(upbeat music) >> Hey everyone. Welcome to theCUBE's presentation of the AWS Startup Showcase. This is Data as Code, The Future of Enterprise Data and Analytics. This is also season two, episode two of our ongoing series with exciting partners from the AWS ecosystem who are here to talk with us about data and analytics. I'm your host, Lisa Martin. Two guests join me, one, a cube alumni. Venkat Venkataramani is here CEO & Co-Founder of Rockset. Good to see you again. And Doug Moore, VP of cloud platforms at Command Alkon. You're here to talk to me about how Command Alkon implemented real time analytics in just days with Rockset. Guys, welcome to the program. >> Thanks for having us. >> Yeah, great to be here. >> Doug, give us a little bit of a overview of Command Alkon, what type of business you are? what your mission is? That good stuff. >> Yeah, great. I'll pref it by saying I've been in this industry for only three years. The 30 years prior I was in financial services. So this was really exciting and eye opening. It actually plays into the story of how we met Rockset. So that's why I wanted to preface that. But Command Alkon is in the business, is in the what's called The Heavy Building Materials Industry. And I had never heard of it until I got here. But if you think about large projects like building buildings, cities, roads anything that requires concrete asphalt or just really big trucks, full of bulky materials that's the heavy building materials industry. So for over 40 years Command Alkon has been the north American leader in providing software to quarries and production facilities to help mine and load these materials and to produce them and then get them to the job site. So that's what our supply chain is, is from the quarry through the development of these materials, then out to the to a heavy building material job site. >> Got it, and now how historically in the past has the movement of construction materials been coordinated? What was that like before you guys came on the scene? >> You'll love this answer. So 'cause, again, it's like a step back in time. When I got here the people told me that we're trying to come up with the platform that there are 27 industries studied globally. And our industry is second to last in terms of automation which meant that literally everything is still being done with paper and a lot of paper. So when one of those, let's say material is developed, concrete asphalt is produced and then needs to get to the job site. They start by creating a five part printed ticket or delivery description that then goes to multiple parties. It ends up getting touched physically over 50 times for every delivery. And to give you some idea what kind of scale it is there are over 330 million of these type deliveries in north America every year. So it's really a lot of favor and a lot of manual work. So that was the state of really where we were. And obviously there are compelling reasons certainly today but even 3, 4, 5 years ago to automate that and digitize it. >> Wow, tremendous potential to go nowhere but up with the amount of paper, the lack of, of automation. So, you guys Command Alkon built a platform, a cloud software construction software platform. Talk to me of about that. Why you built it, what was the compelling event? I mean, I think you've kind of already explained the compelling event of all the paper but give us a little bit more context. >> Yeah. That was the original. And then we'll get into what happened two years ago which has made it even more compelling but essentially with everything on premises there's really in a huge amount of inefficiency. So, people have heard the enormous numbers that it takes to build up a highway or a really large construction project. And a lot of that is tied up in these inefficiencies. So we felt like with our significant presence in this market, that if we could figure out how to automate getting this data into the cloud so that at least the partners in the supply chain could begin sharing information. That's not on paper a little bit closer to real time that we could make has an impact on everything from the timing it takes to do a project to even the amount of carbon dioxide that's admitted, for example from trucks running around and being delayed and not being coordinated well. >> So you built the connect platform you started on Amazon DynamoDB and ran into some performance challenges. Talk to us about the, some of those performance bottlenecks and how you found Venkat and Rockset. >> So from the beginning, we were fortunate, if you start building a cloud three years ago you're you have a lot of opportunity to use some of the what we call more fully managed or serverless offerings from Amazon and all the cloud vendors have them but Amazon is the one we're most familiar with throughout the past 10 years. So we went head first into saying, we're going to do everything we can to not manage infrastructure ourselves. So we can really focus on solving this problem efficiently. And it paid off great. And so we chose dynamo as our primary database and it still was a great decision. We have obviously hundreds of millions of billions of these data points in dynamo. And it's great from a transactional perspective, but at some point you need to get the data back out. And what plays into the story of the beginning when I came here with no background basically in this industry, is that, and as did most of the other people on my team, we weren't really sure what questions were going to be asked of the data. And that's super, super important with a NoSQL database like dynamo. You sort of have to know in advance what those usage patterns are going to be and what people are going to want to get back out of it. And that's what really began to strain us on both performance and just availability of information. >> Got it. Venkat, let's bring you into the conversation. Talk to me about some of the challenges that Doug articulated the, is industry with such little automation so much paper. Are you finding that still out there for in quite a few industries that really have nowhere to go but up? >> I think that's a very good point. We talk about digital transformation 2.0 as like this abstract thing. And then you meet like disruptors and innovators like Doug, and you realize how much impact, it has on the real world. But now it's not just about disrupting, and digitizing all of these records but doing it at a faster pace than ever before, right. I think this is really what digital transformation in the cloud really enable tools you do that, a small team in a, with a very very big mission and responsibility like what Doug team have been, shepherding here. They're able to move very, very, very fast, to be able to kind of accelerate this. And, they're not only on the forefront of digitizing and transforming a very big, paper-heavy kind of process, but real-time analytics and real time reporting is a requirement, right? Nobody's wondering where is my supply chain three days ago? Are my, one of the most important thing in heavy construction is to keep running on a schedule. If you fall behind, there's no way to catch up because there's so many things that falls apart. Now, how do you make sure you don't fall behind, realtime analytics and realtime reporting on how many trucks are supposed to be delivered today? Halfway through the day, are they on track? Are they getting behind? And all of those things is not just able to manage the data but also be able to get reporting and analytics on that is a extremely important aspect of this. So this is like a combination of digital transformation happening in the cloud in realtime and realtime analytics being in the forefront of it. And so we are very, very happy to partner with digital disruptors like Doug and his team to be part of this movement. >> Doug, as Venkat mentioned, access to real time data is a requirement that is just simple truth these days. I'm just curious, compelling event wise was COVID and accelerator? 'Cause we all know of the supply chain challenges that we're all facing in one way or the other, was that part of the compelling event that had you guys go and say, we want to do DynamoDB plus Rockset? >> Yeah, that is a fantastic question. In fact, more so than you can imagine. So anytime you come into an industry and you're going to try to completely change or revolutionize the way it operates it takes a long time to get the message out. Sometimes years, I remember in insurance it took almost 10 years really to get that message out and get great adoption and then COVID came along. And when COVID came along, we all of a sudden had a situation where drivers and the foreman on the job site didn't want to exchange the paperwork. I heard one story of a driver taping the ticket for signature to the foreman on a broomstick and putting it out his windows so that he didn't get too close. It really was that dramatic. And again, this is the early days and no one really has any idea what's happening and we're all working from home. So we launched, we saw that as an opportunity to really help people solve that problem and understand more what this transformation would mean in the long term. So we launched internally what we called Project Lemonade obviously from, make lemonade out of lemons, that's the situation that we were in and we immediately made some enhancements to a mobile app and then launched that to the field. So that basically there's now a digital acceptance capability where the driver can just stay in the vehicle and the foreman can be anywhere, look at the material say it's acceptable for delivery and go from there. So yeah, it made a, it actually immediately caused many of our customers hundreds to begin, to want to push their data to the cloud for that reason just to take advantage of that one capability >> Project lemonade, sounds like it's made a lot of lemonade out of a lot of lemons. Can you comment Doug on kind of the larger trend of real time analytics and logistics? >> Yeah, obviously, and this is something I didn't think about much either not knowing anything about concrete other than it was in my driveway before I got here. And that it's a perishable product and you've got that basically no more than about an hour and a half from the time you mix it, put it in the drum and get it to the job site and pour it. And then the next one has to come behind it. And I remember I, the trend is that we can't really do that on paper anymore and stay on top of what has to be done we'll get into the field. So a foreman, I recall saying that when you're in the field waiting on delivery, that you have people standing around and preparing the site ready to make a pour that two minutes is an eternity. And so, working a real time is all always a controversial word because it means something different to anyone, but that gave it real, a real clarity to mean, what it really meant to have real time analytics and how we are doing and where are my vehicles and how is this job performing today? And I think that a lot of people are still trying to figure out how to do that. And fortunately, we found a great tool set that's allowing us to do that at scale. Thankfully, for Rockset primarily. >> Venkat talk about it from your perspective the larger trend of real time analytics not just in logistics, but in other key industries. >> Yeah. I think we're seeing this across the board. I think, whether, even we see a huge trend even within an enterprise different teams from the marketing team to the support teams to more and more business operations team to the security team, really moving more and more of their use cases from real time. So we see this, the industries that are the innovators and the pioneers here are the ones for whom real times that requirement like Doug and his team here or where, if it is all news, it's no news, it's useless, right? But I think even within, across all industries, whether it is, gaming whether it is, FinTech, Bino related companies, e-learning platforms, so across, ed tech and so many different platforms, there is always this need for business operations. Some, certain aspects certain teams within large organizations to, have to tell me how to win the game and not like, play Monday morning quarterback after the game is over. >> Right, Doug, let's go back at you, I'm curious with connects, have you been able to scale the platform since you integrated with Rockset? Talk to us about some of the outcomes that you've achieved so far? >> Yeah, we have, and of course we knew and we made our database selection with dynamo that it really doesn't have a top end in terms of how much information that we can throw at it. But that's very, very challenging when it comes to using that information from reporting. But we've found the same thing as we've scaled the analytics side with Rockset indexing and searching of that database. So the scale in terms of the number of customers and the amount of data we've been able to take on has been, not been a problem. And honestly, for the first time in my career, I can say that we've always had to add people every time we add a certain number of customers. And that has absolutely not been the case with this platform. >> Well, and I imagine the team that you do have is far more, sorry Venkat, far more strategic and able to focus on bigger projects. >> It, is, and, you've amazed at, I mean Venkat hit on a couple of points that it's in terms of the adoption of analytics. What we found is that we are as big a customer of this analytic engine as our customers are because our marketing team and our sales team are always coming to us. Well how many customers are doing this? How many partners are connected in this way? Which feature flags are turned on the platform? And the way this works is all data that we push into the platform is automatically just indexed and ready for reporting analytics. So we really it's no additional ad of work, to answer these questions, which is really been phenomenal. >> I think the thing I want to add here is the speed at which they were able to build a scalable solution and also how little, operational and administrative overhead that it has cost of their teams, right. I think, this is again, realtime analytics. If you go and ask hundred people, do you want fast analytics on realtime data or slow analytics on scale data, people, no one would say give me slow and scale. So, I think it goes back to again our fundamental pieces that you have to remove all the cost and complexity barriers for realtime analytics to be the new default, right? Today companies try to get away with batch and the pioneers and the innovators are forced to solve, I know, kind of like address some of these realtime analytics challenges. I think with the platforms like the realtime analytics platform, like Rockset, we want to completely flip it on its head. You can do everything in real time. And there may be some extreme situations where you're dealing with like, hundreds of petabytes of data and you just need an analyst to generate like, quarterly reports out of that, go ahead and use some really, really good batch base system but you should be able to get anything, and everything you want without additional cost or complexity, in real time. That is really the vision. That is what we are really enabling here. >> Venkat, I want to also get your perspective and Doug I'd like your perspective on this as well but that is the role of cloud native and serverless technologies in digital disruption. And what do you see there? >> Yeah, I think it's huge. I think, again and again, every customer, and we meet, Command Alkon and Doug and his team is a great example of this where they really want to spend as much time and energies and calories that they have to, help their business, right? Like what, are we accomplishing trying to accomplish as a business? How do we enable, how do we build better products? How do we grow revenue? How do we eliminate risk that is inherent in the business? And that is really where they want to spend all of their energy not trying to like, install some backend software, administer build IDL pipelines and so on and so forth. And so, doing serverless on the compute side of that things like AWS lambda does and what have you. And, it's a very important innovation but that isn't, complete the story or your data stack also have to become serverless. And, that is really the vision with Rockset that your entire realtime analytics stack can be operating and managing. It could be as simple as managing a serverless stack for your compute environments like your APS servers and what have you. And so I think that is going to be a that is for here to stay. This is a path towards simplicity and simplicity scales really, really well, right? Complexity will always be the killer that'll limit, how far you can use this solution and how many problems can you solve with that solution? So, simplicity is a very, very important aspect here. And serverless helps you, deliver that. >> And Doug your thoughts on cloud native and serverless in terms of digital disruption >> Great point, and there are two parts to the scalability part. The second one is the one that's more subtle unless you're in charge of the budget. And that is, with enough effort and enough money that you can make almost any technology scale whether it's multiple copies of it, it may take a long time to get there but you can get there with most technologies but what is least scalable, at least that I as I see that this industry is the people, everybody knows we have a talent shortage and these other ways of getting the real time analytics and scaling infrastructure for compute and database storage, it really takes a highly skilled set of resources. And the more your company grows, the more of those you need. And that is what we really can't find. And that's actually what drove our team in our last industry to even go this way we reached a point where our growth was limited by the people we could find. And so we really wanted to break out of that. So now we had the best of both scalable people because we don't have to scale them and scalable technology. >> Excellent. The best of both worlds. Isn't it great when those two things come together? Gentlemen, thank you so much for joining me on "theCUBE" today. Talking about what Rockset and Command Alkon are doing together better together what you're enabling from a supply chain digitization perspective. We appreciate your insights. >> Great. Thank you. >> Thanks, Lisa. Thanks for having us. >> My pleasure. For Doug Moore and Venkat Venkatramani, I'm Lisa Martin. Keep it right here for more coverage of "theCUBE", your leader in high tech event coverage. (upbeat music)

>> (narrator) From around the globe. It's theCUBE. With digital coverage of AWS re:Invent 2020. Sponsored by Intel, AWS and our community partners. >> Welcome to theCUBE virtual. Our coverage of AWS re:Invent 2020 continues. I'm Lisa Martin. Got a couple of guests joining me next. Wendy Moore the VP of product marketing from Trend Micro is here and Geva Solomonovich Global Alliances CTO from Snyk. Wendy and Geva, It's great to have you both on the program today. >> Thanks for having us. Great to be here. >> Hi, thanks for having us. >> Last year we were probably all crammed in Vegas together. Here we are virtually but it's great that we're still able to connect. So lot has gone on since we were all at re:Invent in Vegas last year. Wendy, let's start with you from a security perspective there's been a growth in open source vulnerabilities that have impacted enterprises globally. Talk to me about what you're seeing there. What's going on? >> Yeah. Well. I think everybody in this audience recognizes the rapid shift to the use of open source in development teams. And what we've seen alongside that is a rapid increase in the number of vulnerabilities that are showing up in open source software. So that means that vulnerabilities that can be exploited and cause damage to your company's application, reputation and your customers, are on the increase out there. >> And a number that you sent over was two and a half X growth in open source vulnerabilities in the last year. Has that number gone up during the pandemic? >> So I'm not sure if the vulnerabilities have gone up during the pandemic, but we've definitely seen an increase in exploitation of vulnerabilities. There's so much in the news about ransomware incidents in healthcare targeting pharmaceutical organizations, and most of those are taking advantage of vulnerabilities. Not necessarily in open source, but some of it is definitely happening in open source. >> Now we've been talking about the rise in ransomware for awhile, and it's all... The numbers and types of companies and healthcare organizations like is it schools, governments, for example lot of vulnerabilities being exploited that's for sure. >> So Geva let's go over to you. Talk about from Synk's perspective. The impact on businesses and how can you guys help. >> And then I'll put in a few insights there. on the open source risk. Wendy talked about it as well. Why is it growing? One of course is open source tuition usage is growing. So of course it bulges, the amounts of vulnerabilities is growing and the amount of exploits. But when you look at it from a hacker's perspective, attacking is an ROI based activity. Hackers want to spend their hacking hours where they're more likely to get our reward, be able to get that ransom or steal the data or do whatever they can. And open source actually makes it much easier for them than a lot of these other alternatives. One, the source is open. So just finding a vulnerability is much easier than trying to find the vulnerability in proprietary code. Two, there's like a market for these exploits and companies even like need for chapter. One of the byproducts of that is you can just go and feel the vulnerabilities out there and pick the ones that you want to try to exploit. But three, which is really the most critical piece is that if you do find the juicy vulnerability in a very popular open source package, the amount of companies you can attack is not one, is thousands or tens of thousands because that's precisely what makes the popular open source packages popular. It's being used broadly and so if you spend this effort to develop an exploit and then you can send it like there just across the world to 10 thousands of companies you're more likely to be successful. And that's what's driving a lot of the hacker attention into the open source vulnerabilities and that's why the growing. >> So it's a low cost high reward for those hackers. Wendy what are some of the ways that organizations can protect themselves from this? >> Well, one of the best ways to protect themselves against exploitation of vulnerabilities and against vulnerability showing up in their code is to actually analyze their code and scan it looking for vulnerabilities. And the best possible place to do that is actually in the code repository. So before code is ever packaged up and deployed it actually gets caught really early. So it's all about shifting security left. But some of the challenges with that is that you know the code repository, Tory and the code and open source has largely been the domain of DevOps and the developers and security who is tasked with managing the risk of the organization has little to know visibility into what vulnerabilities might exist. So something that's a growing part of an enterprise risk profile the security team doesn't really see. And that's a big gap for most organizations. >> So in terms of that visibility being essential, sounds like maybe even a cultural gap there. Geva what are your recommendations? We, you know, we talk about SecOps, we talk about DevOps. Is the solution DevSecOps or SecDevOps? >> I mean, all these partners are definitely helping there but you kind of need to break it down and understand what their problems, which is what Wendy was articulating. Why you have these traditional security teams have all their traditional tools. They look at mostly and let's call it the IC type security. Then you have this entire new category of risk which is lets say open source risk, but it's just inside the code repository inside a GitHub repo or somewhere, or they completely have no visibility into. And what that causes is one has to have a conversation with the developers who are those who are convenient to pick those vulnerabilities, remove them from the code. And, but to also, just from the mind ensuring that in our location it's hard for you to protect something that you don't have visibility into which causes opensource security to be possibly under provisioned in your entire a security fence. As you're looking at the security risk. And as we are talking about solution, so one of the movements we've seen with DevOps, where you know engineering team and IT teams have come together to have a shared ownership of the results of deploying these applications. In production now you expand out into DevSecOps. It's okay to actually make this work. We need to have a shared responsibility model where both developers step up to take some ownership and the traditional security each step up to understand what the developers are doing, build tools to make it easier for them. And ultimately I think Wendy nailed it on the head. She said the best way to protect yourself is actually to remove the vulnerable line of code from your application, not wait for it to be deployed and try to put some blocks in there. >> All right. So Wendy how are Trend Micro and Snyk working together to resolve that challenge that you guys just described? >> Yeah, we'll Trend Micro and Snyk have been working together for over a year now. And we came out with an initial offering and now we're coming out with a new offering that is really focused on basically delivering that code scanning ability right in the code repository. And through Trend Micro's Cloud One platform, we are delivering this as a service to the security operations team so that they get visibility of anything that Snyk finds in the code repository. And they can take action from there. So Trend Micro's Cloud One security services platform basically equips cloud builders with a whole bunch of different types of technologies to satisfy their different infrastructure requirements. So we've got things like workload security application security, network security, a number of different take types of security tools. And this just brings another security tool to the security operations team and the DevOps team so that they can basically extend their visibility and their security controls back to the code repository. >> Geva what are some of the impacts that you're seeing. So for obviously besides wanting to find those vulnerabilities faster as when you talk about shifting left. Give me some examples of some customers that you were working with maybe in the first iteration and what the impact has been. >> The impact is the... what, sorry, can you repeat the question? >> Yeah. Impact of your technologies together? You said that there's a new offering coming up but talk to me about some of the impact that these customers are making. >> Yeah. Okay. Sorry. Thank you for repeating the question. And so this joint product is very cunning from a multiple perspective. So one, it's going to be delivered inside the Cloud One platform, which Wendy just talked about. You asked before what is the impact of COVID? And one of the big impacts has been on the financial stress. Every company in every, every vendor is having. And so just the ease of managing less vendors and less tools and less places to procurement is of high value for every organization Just in terms of efficiency of operations. And just being able to acquire this new product on an existing platform where there are already consuming security tools. That by itself is amazing value. And number two, we're taking again... We're taking a technology which is a cloud native, it's a modern technology. And that's typically has been outside of the purview of a traditional security team and making it accessible to them in a place where it's easy for them to try out and they can, you know, start small and grow from there. They don't have to make a big commitment to get going. And more importantly, it's giving them visibility into this important technology that they didn't have before. >> So Wendy this is all intended at bridging that gap? I'm just curious, like if we take a peek inside, what this enables SecOps to do what it enables DevOps to do. What were some of the feedback that you're hearing from customers about those teams coming together and actually being able to work very collaboratively with that shift left actually being able to be done? >> Yeah. I mean, you know, if you talk to... There's some organizations who do this really well. They're very mature and their security operations teams and their DevOps teams work very closely together collaboratively, excuse me. And they also understand each other's needs. So they're able to insert tools into the security pipeline that don't slow DevOps down but also meet the needs of the security team. Whereas we see some other organizations where Dev is at one side of the pipeline and you've got security at the other and they don't tend to converse or meet. And those are the organizations where there tends to be more challenges. So the idea with this new solution is it's going to give the security team visibility of basically the scale and scope of their open source situation. So that they've actually got some data to go have conversations with the DevOps teams and start going in that direction of making those teams work more seamlessly together. I mean, you used the term DevSecOps before, some organizations that's a very real situation. Others still have a long way to go. And we think this is a great first step to bring those teams together. >> Fostering long-term friendships I'm sure. Just talk to me about the go to market, Wendy. How are you guys going to market together? Trend Micro and Snyk selling direct channel? What is it like? >> So this is actually going to be a Trend Micro Cloud One offering. So we jointly developed it with Snyk but it's going to be Trend Micro who is selling it. And we go to market a number of different ways. AWS marketplace is a big channel to market for us And this will be available for purchase there. When it becomes available in January. And also, we also work very closely with channel partners as well who also participate in AWS marketplace. >> So what are some of the things that you're expecting to customers to be able to take advantage of around the time of re:Invent and into early 2021? >> Yeah. I really encourage customers to visit our page on the AWS re:Invent platform. We're going to have all kinds of exciting demos there. You can go learn more about this new offering that we're delivering jointly developed with Snyk. And you can also ask about how you can sign up for early access to this new offering. So highly encourage you to go check that out. >> Excellent, early access is always nice to be a beta tester and really get that symbiotic relationship. >> Geva last question for you is as the Global Alliances CTO I imagine your customer conversations in the last year have changed dramatically. Talk to me about some of the things that you really think like in terms of like exposing vulnerabilities. Let's talk about exposing opportunities that that Snyk is helping organizations do so that they can not just keep the lights on during this very unprecedented time but actually be winners of tomorrow. >> Yeah, I think again at the heart of the DevOps movement and why it's been successful it's reducing that feedback loop between writing some codes, getting it to production in the hands of customers, getting the feedback from them and rinse and repeat and starting that loop. And those who have it, the faster you can get to market faster and can deliver value faster ultimately are the winners. Now, one of the things we've seen with the COVID is a lot of the this outbound activity has been going down. People have been going less to events and need to look more internally and how you can become better as an organization. And you've actually seen an increase in the investment of a digital transformation and cloud journeys and stuff like that. And one of the... One of kind of the traditional inhibitors that's going fast and all in into the cloud is the loss of control of the traditional security teams on the application development. Where now people can, you know... deploy hundreds of times every application to the cloud a day. And what we've seen is that they come to Snyk or to companies like ours, so we can secure those new modern development life cycles and give the security feedback to the developers as they're building the applications and give the security teams the visibility into those pipelines and application domain. So they have a sense that they're not losing all the control they used to have. They're still getting visibility into those application development and actually allowing their organizations to go faster because of it they can sign up to and be doing the technologies and actually increase the speed of going to the cloud. >> Yeah and that's critical because as we, you mentioned as we've been talking about for months now that the acceleration of cloud adoption, the speed of digital transformation it's one of those things that's challenging to do. You've got to have visibility. Period. In order to facilitate that. And if it's another thing that you kind of were describing Geva as that visibility provides that sense of control or trust, and that's also huge for not just a business to catch vulnerabilities but for teams the DevOps teams, the SecOps teams to be working together in a highly collaborative way. Do you agree Wendy? >> Absolutely. And the beautiful thing is this sets that up This tool. So it allows them to work together very collaboratively but it also sets up that visibility. So that down the road there could be even further automation into that process. Because you know, the whole purpose of DevOps is to take the people out of it. Right. So, but in order... You need to set up those processes to begin with. So this is a first step in terms of setting up that automation and visibility amongst those two teams. >> Excellent. And can you say one more time Wendy where prospective customers can go to learn more and become a early adopter? >> Yeah, absolutely. So visit our Trend Micro page at the AWS reinvent platform. And there you'll be able to learn much more about the offering and also learn how you can access the early adopter program. >> Excellent. You guys thank you so much for joining me on the program today. Sharing what Trend Micro and Snyk are doing together and how you're helping organizations cross-functionally be successful. We appreciate your time. >> Thank you, Lisa. Appreciate it. >> Thank you so much. >> My pleasure. For my guests, I'm Lisa Martin and you're watching theCUBE virtual. (upbeat music)

>> Introducer: From the CUBE studios in Palo Alto and in Boston connecting with thought leaders all around the world. This is a CUBE Conversation. >> Hi everybody this is Dave Volante. Welcome to the special CUBE Conversation. I'm really excited to invite in my mentor and friend. We go way back. Fred Moore is here. He's the president of Horizon Information Strategies. We going to talk about managing data in the zettabyte era. Fred, I think when we first met, we were talking about like the megabyte era. >> Right, exactly. I think back then we had, you know, maybe 10 bytes in our telephone and one on the wristwatch, you know, but now you can put a whole data center in a single cartridge of tape and take off. Things that really changed. >> It's pretty amazing. And of course, for those who don't know Fred, he was the first a systems engineer at Storage Tech. And as I said, somebody who taught me a lot in my early days, of course he's very famous for the term that everybody uses today. Backup is one thing, recovery is everything. And Fred just wrote, you know, this fantastic paper. He's done this year after year after year. He's just dug in, he's a clear thinker, strategic planner with a technical bent in a business bent. You're like one of those five tool baseball players, Fred. But tell me about this paper. Why, did you write it? >> Well, the reason I wrote that is there's been so much focus in the last year or so on the archive component of the storage hierarchy. And the thing that's happening, we're generating data lots faster than we're analyzing it. So it's piling up being unanalyzed and sitting basically untapped for years at a time. So that has posed a big challenge for people. The other thing that got me deeper into this last year was the Hyperscale market. They are, those people are so big in terms of footprint and infrastructure that they can no longer keep everything on disk. It's just economically not possible. The energy consumption per disk, the infrastructure costs, the frequency of, you know, taking a disc out every three, four or five years for just for replacement, has made it very difficult to do that. So Hyperscale has gone to tape in a big way, and it's kind of where most of the tape business in the future is going to wind up in these Hyperscale businesses. >> Right. >> We know tape doesn't exist in the home. It doesn't exist in a small data center. It's only a large scale data center technology, but that whole cosmos led me into the archive space and in a need for a new archive technology beyond tape. >> So, I want to set up the premise here. Just going to pull this out of your paper. It says a 60% of all data is archival, and could reach 80% or more by 2024, making archival data by far the largest storage class. And given this trajectory, the traditional storage hierarchy paradigm is going to to need to disrupt itself. And quickly we're going to talk about that. That really is the premise of your paper here, isn't it? >> It is, you know, to do all this with traditional technologies is going to get very painful for a variety of reasons. So the stage is set for a new tier and a new technology to appear in the next five years. Fortunately, I'm actually working with somebody who is after this in a big way, and in a different way than what you and I know. So I think there is some hope here that we can redefine and really add a new tier down at the bottom. You see it kind of emerging on that picture of the deep archive tier it's. Beginning to show up now and it's, you know, infinite storage. I mean, if you look at major league sports, the world series and Superbowl, you know, that data will never be deleted. It'll be here forever. It'll be used periodically based on circumstances. >> Yeah, well, we've got that pyramid chart up here. I mean, you invented this chart, essentially. At least you were the first person that ever showed it to me. I honestly think that you first created this concept where you had a high performance tier, and a high cost per bit, and then an archive tier. Maybe it wasn't this granular, you know, back in the '70s and '80s? But it's constantly been changing with different media types and different use cases. >> You know, you're right. I mean, and you all know this because you know, when storage deck introduced the nearline architecture, nearline set in between online and offline storage, we called it nearline, and trademarked that term. So that was the tape library concept to move data from offline status to online status, with a robotic library. So that brought up that third tier online, nearline, and offline, but you're right. This pyramid has evolved and morphed into several things. And, you know, I keep it alive. Somebody said, I'll have a pyramid on my tombstone instead of my name when I go down. (both chuckles) But it's really the heart and soul of the infrastructure for data. And then out of this comes all the management and security, the deletion, the immutable storage concepts, the whole thing starts here. So it's like your house, you got to have a foundation, then you can build everything on top of it. >> Well, and as you pointed out in your paper, a minute ago, it always comes down to economics. So I want to bring up the sort of 10 year expected cost of ownership the TCO for the three levels you got all disk, you got all cloud and you got LTO and you got the different aspects of the cost. The purple is always the biggest piece of cost. It's the labor costs. But of course, you know, in cloud, you've got the big media cost because they've done so much automation. I wonder if you could take us through this slide, what are the key takeaways there? >> Well, you know the thing that hurts here with all these technologies is, as you can see up on top up there, what the key issues are with this and the staff and personnel. So the less people you have to manage data, the better off you are. And then, you know, it's pretty high for disk compared to a lot of things to do on desk, but lack of manage a lot of, you know, sadly what you and I had to deal with years ago and provision kind of, I mean, a lot of this stuff is just labor intensive. The further you get, the further down the pyramid and you also get less labor intensive storage. And that helps then you get a lower cost for energy and cost of ownership. The TCO thing is kind of taking on a new meaning. I hate to put up a TCO chart in some regards, because it's all based on what your input variables are. So you can decide something different, but we've tried to normalize all kinds of pricing and come up with everything. And the cloud is a big question for most people as to how does it stack up. And if you don't ever touch the data in the cloud, you know, the price comes way down. If you want to start moving data in and out of the cloud, you're going to have to ante up in a big way like that. But, you know we're going to see dollar a terabyte storage prices down at the bottom of this pyramid here in the next five years. But hey, you can get down to four or five terabyte with drives media in libraries tape, just entire flash and certainly higher than that. But you know, we're going to have the race to a dollar a terabyte, total TCO cost here in 2025. >> So when Amazon announced, they just announced a glacier. Everybody said, okay, what is that? Is that tape is that, you know, this spun down disk, cause it took a while to get it back. But you're kind of seeing that tape technology as you said, really move into the Hyperscale space and that's going to accommodate this massive, you know, lower part of the pyramid, isn't it? >> Exactly. Yeah. And we don't have a spin down disk solution today. I was actually on the board of a company that started that called Copay and years ago, right up here near Boulder. >> You watch him (both chuckles) You absolutely right. And a few other people that, you know also, but the spin down disk never made it. And you know, you can spin up and down on a desk on your desktop computer, but doing that in a data center, then on a fiber channel drive never made it. So we don't have a spin down disk to do that. The archive space is kind of dominated by very high capacity disc and then tape. And most of the archive data in the world today, unfortunately sits on display. It's not used and spinning seven by 24, three 65 and not touch much. So that's a bad economic move, but customers just found that easier to handle by doing that then going back to tape. So we've got a lot of data stored in the wrong place from a total economics point of view. >> But the Hyperscalers are solving this problem, or they're not through automation. And, you know, you referenced storage, tiering, really trying to take the labor cost out. How are they doing? Are they doing a good job? >> They've done really well taking the labor costs down, I mean, they have optimized every screw, nut and bolt in the 42 chassis that you could imagine to make it as clean as possible to do that. So they've done a whole lot to bring that cost down, but still the magnitude of these data centers, we're going to finish the year 2020 with about 570 Hyperscale data centers. So it's going right now around the world. You know, each one of these things is 350 400,000 square feet, and up of race wars space. And the economics just don't allow you to keep putting inactive data on spinning disk. We don't have to spin down disk, tape You know, I feel like the only guy in the industry that says this sometimes, but, you know, tapes had a, you know, a renaissance. That people don't appreciate in terms of reliability, throughput, you know, tapes three orders of reliability higher than disc right now. And most people don't know this. So tape's viable, the Hyperscalers see that. And read one Hyperscalers or you know, by over a million pieces of LTO tape last year alone. Just to handle this, you know, be the pressure valve to take all of this inactive stuff off of the gigantic disc farms that they have. >> Well, so let's talk about that a little bit. So you just try to keep it simple. You've got, you know, flash disk and tape. It feels like disc is getting squeezed. We know what flash has done in terms of eating into disc. And you see in that, in the storage market generally, it's soft right now. And I've posited that a lot of that is the headroom that data centers have with flash, is they don't have to buy spindles anymore for performance reasons. And the market is soft. Only pure is showing consistent growth, and ends up a little bit, cause because of mainframe, you've got Dell popping back and forth, but generally speaking, the primary storage market is not a great place to be right now, all the actions and sort of secondary storage and data protection. And so just going to get squeezed, and you mentioned tape, you said that if your only person talking about it, but you said in your paper, you know, it's sequential. So time to first bite is, is sometimes problematic, but you can front end a tape with cash. You can use algorithms and, you know, smart scans and to really address that problem. And dramatically lower the cost. Plus you could do things like you tell me Fred, you're the technologists here, but you're going to have multiple heads things that you can't necessarily do in a hermetically sealed disc drive. >> (chuckles) You can. And what you just described is called the active archive layer in the pyramid. So when you front end a tape library with a disk array for a cash buffer, you create an active archive and that data will sit in there three or four or five days before it gets demoted based on inactivity. So, you know for repetitive use and you're going to get dislike performance for tape data, and that's the same cash in concept that deserve systems had 30 years ago. So that does work and the active archive has got a lot of momentum right now. There's right here near me, where I live in Boulder. We have the Active Archive Alliances headquarters, and I get to do their annual report every year. And this whole active archives thing is a big way to make and overcome that time, the first bike problem that we've had in tape. And we'll have for quite a while. >> In your paper, you've talked about some of the use cases and workloads and you laid out, you know basically taking the pyramid and saying, okay based on the workload, some certain percentage should be up at the top of the pyramid for the high performance stuff. And of course lower for the, you know, the less, you know, important traditional workloads, et cetera. And it was striking to see the Delta between annual, the highest performance we had 70% , I think was up in the top of the pyramid versus, you know the last use case. So in you're talking about what it costs to store a zettabyte in services is that if I talk about 108 million at the high end versus a about 11 or 12 million, so huge Delta 10 X Delta between the top and the bottom based on those, you know allocations based on the workload. >> Yeah, I tried to get at the value of tiered storage based on your individual workload in your business. So I looked at five different workloads, the top one that you referenced. That was in there at 108 million, you know, is the HPC market. I mean, when I visited a few of the HPC people, you know, their DOD agencies in many cases, you know that and I threw the pyramid up. The first thing they would say our permanents inverted. You know (chuckles), all of our archive data is about 10%. You know, we were all flash as much as we can. And we have a little bit archived, we're in constant. Simulation and compute mode and producing results like crazy from the data. So we do an IO, bring in maybe a whole file at a time and compute for minutes before we come up with an answer. So just the reverse. And then I got to look into all the different workloads talking to people, and that's how we develop these profiles. >> So let's pull up this future of the storage hierarchy, was again kind of of talks to the premise of your paper. Walk us through this like, what changes should we be expecting, and you got air gap in here. We're going to, I'm going to ask you about remastering and lifespan, but take us through this. >> Yeah, you know, the traditional chart that you had up on the first big year had four tiers, you know, two disturbs and solid state at the top. And then the big archive tier, which is kind of everything falling down into tape at this point. But you know again, tape has some challenges. You know time to first bite and sequential access on. And then when we couple using tape or disc as an archive, most of that data that's archival is captured as unstructured data. So we don't have, we don't have tags, we don't have metadata, we don't have indices, and that has led to the movement for object storage, to be a primary, maybe in the next five years, the primary format in store archived data, because it's got all that information inside of it. So now we have a way to search things and we can get to objects, but in the interim, you know, it's hard to find and search out things that are unstructured and, you know, most estimates would say 80% of the world's data is at least that much is unstructured. So archives are hard to find once you store it, there's one storing is one thing, retrieving it is another thing. And that's led to the formation of another layer in the story tier. It's going to be data that doesn't have to be remastered or converted to a new technology. in the case of the disc, every three, four or five years or tape drive every eight, maybe 10 years take large lost. Kate Media can go 30 years, but with all new modern tape media, but unfortunately, you know, the underlying drive doesn't go back that far, you can't support that many different versions. So the media life is actually longer than it needs to be. So the stage is set for a new technology to appear down here to deal with this archives. So it'll have faster access will not need to be remastered every five or 10 years, but you'll have, you know, a 50 year life in here. And I believe me, I've been looking for a long time to be able find something like this. And, you know we have a shot at this now, and I'm actually working with the technology that could pull this off. >> Well, it's interesting also as well, you calling out the air gap and the chart we go back to our mainframe guesses, is not a lot we haven't seen before, you know, maybe data D duplication, but you know, the adversary has become a lot more sophisticated. And so air gaps and, you know, ransomware on everybody's mind today, but you've sort of highlighted three layers of the pyramid that are actually candidates for that air gapping. >> Yeah. The active archive up there, of course, you know, with the disk and tape combined, then just pure tape. And then this new technology, which can be removable. You know, when you have removability you create an air gap. little did we know when you and I met that removability would be important to take. We thought we were trying to get rid of the Chevy truck access method, and now without electricity with a terrorist attack and pandemic or whatever. The fastest way to move data is put it on a truck and get it out of town. So that has got renewed life right now. Removability much to my shock from where we started. >> You talked about remastering and you said it's a costly labor intensive process that typically migrates previously archived data to new media every five to 10 years. First of all, explain why you have to do that and how a data center operators can solve that problem. >> Yeah. And let's start with data where most of it sits today on described, you know it describes useful life is four to five years before it either fails or is replaced. That's pretty much common now. So then they have to start replacing these things. And that means you have to copy, you know, read the data off the disk and write it somewhere else, big data move. And as the years go by that amount of data to revamp or gets bigger and bigger. So, I mean, you can do the math as you well know, you want to move, you know, 50 petabytes of data. It's going to take several weeks to do that electronically. So this gets to be a real time consuming effort. So most data centers that I've seen will keep about one fifth of their disposal every year migrating to a new technology, just kind of rolling forward as they go like that rather than do the whole thing every five years. So that's the new build in the disc world. And then for tape the drive stay in there longer, you know the LTO family drives a good read. You know two generations back from the current one that's been there. They cut that off a year ago. They'll go back to something like this soon. But you know, you can go into 10 years on a tape drive. The media life because of very unfair right media, which was already oxidized the last 30 years or more. The old media metal particle was not oxidized. So, you know, the oxidized flake, the particles would fall off people will say shit. I've had this in here eight years, you know, and it's kind flake it I put it back in. So that didn't work well. But now that we had various Verite Media, it was all oxidized, the media lives skyrocket. So that was the whole trick with tape to get into something that was preoxidized before time could cause it to decay. So the remastering is a lot, is less on tape by two to one to three to one, but still when you've got petabytes, maybe an exabyte sitting on tape in the future, that's going to take a long time to do that. >> Right. >> So remastering you'd love a way to scale capacity without having to continue to move the data to something new ever so often. >> So my last question is you've , you know, you went from a technical role into a strategic planning role, which of course the more technical you are in that role, the better off you're going to be. You don't understand that the guardrails, but you've always had a sort of telescope in the industry and you close the paper and it's kind of where I want to end here on, you know, what's ahead. And you talk about some of the technologies that obviously have legs, like three D NAND and obviously magnetic storage. You got optical in here, but then you've got all these other ones that you even mentioned, you know, don't hold your breath waiting for these multilayer photonics and dedic DNA. What class media, holographic storage, quantum storage we do a lot about quantum. What should we be thinking about and expecting as observers as to, you know, new technologies that might drive some innovation in the storage business? >> Well, I've listed the ones that are in the lab that have any life at all, right on this paper. So, you know can kind of take your pick at what goes on there. I mean, optical disk has not made it in the data center. We talked about it for 35 years. We invested in it in storage deck and never saw the light of day. You know, optical disk has remained an entertainment technology throughout the last 35 years. And the bigger rate is very low compared to data center technology. So, you know optical would have to take a huge step going forward. We got a lot of legs left in the solid state business. That's really active SSB, the whole nonvolatile memory spaces. Probably not 45% of the total disc shipments in terms of units, from what it was at it's high and in 2010. Unbelievable though. You know, in disc shipment 650 million drives a year announced just under 400, 35,400. So flashes has taken this stuff away, like crazy. Tape shouldn't be taking just away, but the tape industry doesn't do a very effective job of marketing itself. Most people still don't know what's going on with tape. They're still looking out of the roof, still looking out of the rear view mirror at a tape, as opposed to the front windshield. We see all the new things that have happened. So, you know they have bad memories of taping the past load stretch, edge damage tape, wouldn't work a tear or anything like that. It was a problem. Oh, that's pretty well gone away now. In a moderate tape is a whole different ball game, but most people don't know that. So, you know tapes going to have to struggle with access time and sequential reality. They've done a few things to come over excess time and the order request now to take the optimizer based on physical movement on the tape that can take out 50% of your access time for multiple requests on a cartridge. The one on here that's got the most promise right now would be a version of a multilayer photonic storage, which is. I would say sort like optical, but, you know, with data center, class characteristics, multi-layer recording capability on that random access, which tape doesn't have. And, you know, I would say that's probably the one that you would want to take some look at going forward like this. The others are highly specular. You know, we've been talking about DNA since we were kids. So we don't have a DNA product out here yet. You know, it's access times eight hours. It's probably not going to work for us. That's your, that's not your deep archive anymore. That's your time capsule storage. >> Yeah, right. >> Lock the earth. So, I mean, I think you kind of see what's here. I mean, the chances are it's still going to be the magnetic technologies tape disc, and then the solid state number and stuff. >> Right. >> But these are the ones that I'm tracking and looking at, trying to have worked with a few of the companies that are in this. Future list and I'd love to see something breakthrough out there, but it's like, we've always said about a holographic storage. For example, you know, there's been more written about it than there's ever been written on it. (both chuckles) >> Well, the paper's called Reinventing Archival Storage. You can get it on your website I presume Fredhorizon.com >> Yep, absolutely. >> Awesome. >> Fred Moore, great to see you again. Thanks so much for coming on the CUBE. >> My pleasure, Dave. Thanks a lot. Great job. >> All right. And thank you for watching everybody. This is Dave Volante for the CUBE. We'll see you next time. (upbeat music)

>> Live from Las Vegas It's theCUBE covering the AWS Accenture Executive Summit. Brought to you by Accenture. >> Welcome back everyone to theCUBE's live coverage of the AWS Executive Summit here at the Venetian in Las Vegas. I'm your host, Rebecca Knight. We have two guests for this segment. We have Mike Moore, Senior Principal at Accenture Research, and Andrew Wilson, Chief Information Officer at Accenture. Thank you both so much for returning to theCUBE. >> Good to see you as ever, Rebecca, and to be back in Las Vegas as well. >> Exactly, back in Sin City, right, here we are. So our topic is innovation. A buzzword that is so buzzy it's almost boring. Let's start the conversation with just defining innovation. What does innovation mean? >> An objective, a behavior, a way of working. To me, innovation is what we need to do with modern technology to enable the enterprise and the business world and be creative humans and to use disciplines which we didn't typically bring to work before. >> And is it creativity, or is there sort of logic and rationale too? >> I think there's logic and rationale. But there's also entertainment, fun, modern consumer-like experimentation, risk-taking, things of that nature. >> I think that a big key is actually striking a balance between creativity and logic and rationale and that's the really tricky bit, because you need to give your employees the license to be creative but within a certain set of boundaries as well. >> The rules of work have definitely changed, and behaviors that we encourage, even the clothes we wear, how we work, when we work, those are all characteristic of a more innovative, accepting diverse world, and a world that can keep up with the modern technology and the advancements and the announcements like we're hearing about here at re:Invent. >> It's the ultimate right brain, left brain behavior and activity. So Mike, you've done some research recently about the hallmarks of innovative companies, what they do differently from the ones that are not innovative, that are failing here, so tell our viewers a little bit about what you've found in your research. >> We surveyed 840 executives from a variety of different companies, different industries, different geographies, to understand their approach to innovation, and those who were doing it particularly well, and those maybe not so well. And around about 14 percent of our respondents were turning their investments in innovation into accelerated growth, and there were lots of different reasons for their success but three things really stood out. So first of all their outcome lacked in terms of the way they approach innovation, so they put a clear set of processes around their innovation activities, and then linked those to operational and financial performance metrics. They're also disruption minded, so they're not just pursuing incremental tweaks to their products and services, but their investing in disruptive technologies that could actually create entirely new markets. And then finally they're change orientated. They're not just using innovation to change their products and services, but also to fundamentally change the nature of their own organizations as a whole. >> So 14 percent are knocking it out of the park. Does that mean the rest of them are all laggards or are sort of some in the middle? What is the state of innovation in industry today, would you say, Andrew? >> I would say it's hugely variable by industry, geography, type of company, and individual instance of leader and culture, but I am sure that the most successful companies, those that are pivoting to the new, those that are imaginative, those that have recently arrived, all have that DNA that we're describing, all have that way of working, all have that ability to operate cleverly, intelligently, humorously, and at speed. I think innovation is very much characterized by something that can be fast-failed, do, step, move sideways, do again. The way of working has changed in modern enterprises. We as CIO's have to accept that. We have to speed up. We have to create the environment in where that productivity, where that creation can occur, and I think all of that's key. >> You keep mentioning this, the way of working has changed, and I think we all sort of know what you mean but explain a little bit what you're seeing. >> Experimentation, the ability to get more done with the resources that you have. So here we are at AWS re:Invent, cloud-based operations. Cloud gives you, gives me as a CIO the means to do more, more quickly, more rapidly, on a greater scale, in more places that I ever could have imagined in my old old-fashioned data senses. So the services we can consume, the data we can connect together, the artificial intelligence we can bring to it, the consumer-like experience. All of those things, which by the way, are drawing on innovative behaviors in their own right, are absolutely what the game is about now. >> How does AWS figure into your cloud transformation? >> Well for our cloud transformation at Accenture, AWS is one of the core cloud platform providers who power Accenture. We are nearly 95 percent in cloud. So as an organization that's very pronounced, and typically ahead of most organizations. But we sort of have to be, don't we? I mean, we have to be our own North Star. I can't sit here and explain the virtues of what Accenture can bring to a client's cloud transformation if we haven't already done it to ourselves. And by the way, that drew on innovative approaches, risk-taking approaches because over the last three years we've moved Accenture to the cloud. >> So I love how you said it, we are our own North Star, and other people would say we eat our own dog food, I mean that's just kind of more gross, but in terms of having experienced this transformation yourselves, how do you use what you've learned to help your companies transform as well? And make these moves, take these risks, what would you say to that? >> Well I think we keep an eye on the research with our colleagues there, they're our own North Star. I think we look at the ecosystem, we assess readiness for enterprise, security compliance, scale, availability, and then we also look and say, and what's ready for prime time in terms of Accenture scale, half a million people nearly. You bring all of those things together and it's a recipe, and that's why we consult our business, that's why we guide and educate and experiment and innovate together. And that's very much how we adopted cloud, it's very much how we do a number of other things, and the creative services we have. >> In terms of, let's get back to the research. So how do you, I mean as you said, the research is, as Andrew said, it's something that executive leaders are looking at to figure out what's actually happening in the market as well as what's happening within the organization itself. So how do you set your research agenda in terms of figuring out where you want to focus your time and energy and resources. >> Well I think we do it in a very similar way to in which we consult with clients, we speak to them. We talk to them about some of the key issues that they're facing and we always interview a series of executives and also academics to get their perspective at the start of their project. And that's something that we did in this particular instance and what we heard from many executives was that, to the point that Andrew was making before, the speed and scale of innovation today is happening at a completely different pace than in the past. So product cycle times are just diminishing in every single industry and as a consequence, executives now need to build new innovation units to make sure that they can respond to that changing market. So that's we wanted to explore through the research. >> So in this research, with the 14 percent doing it well, the 86 percent sort of either, somewhere on the spectrum of doing terribly or figuring things out, getting better, what are their pain points, and what's your advice to those companies? >> Well I think, and we take the positive spin on it in terms of what the companies are doing well, one of the points that Andrew was making before was how Accenture works with other partners to become more innovative itself. And that's something that we saw many of the high performing companies doing. So many of them were what we call networks powers. Not just innovating using their own resources, their own people, but their drawing on a broader ecosystem of partners to bring the very best products and services to their customers, and their spending not just on R and D internally but also on accelerators, incubators, technology based M and A, and actually their spending as much on inorganic innovation as they are on organic innovation. >> At Accenture we actually help our clients look for trap value, and what we mean by that is if an organization with a history, with a set of business processes, a set of technologies, and a set of disciplines and employees that have been successful and worked possibly for decades in that model, then they're going to be in some pretty tight guide rails. How do you innovate out of that, to deal with all of the destruction that's now available, good healthy disruption, that actually reveals the next level of efficiency, customer satisfaction, product creativity, and innovation in it's own right, so that's innovation in action, if you like. >> I want to ask, here we are at AWS re:Invent, Andy Jassy on the main stage this morning announcing a dizzying number of new products, services, and AWS, this is Amazon, this is a huge company that really seems to know how to innovate, and do it constantly, but is that is that, can every company be Amazon? You know what I'm saying? I mean, is this really possible and attainable? >> Is such a thing as innovation fatigue perhaps? >> Well, exactly, right! >> My view is that you have to find a way to make innovation a constant and a norm. It doesn't mean that you always will have to operate with the same ridiculous pace, but creativity and pace do go hand in hand to a point, but to be ahead, to stay ahead, and to lead an organization of technologists, who can comprehend all of these announcements, so you have to innovate in both how you lead and operate as well. It's not just your product, it's your behaviors, because there's just so much coming all the time. >> Right, and we've seen a number of large companies, not necessarily technology companies, but I'm thinking of Sears and Toys-R-Us, that have really, you've seen what can happen, the cautionary tales. >> Look at the attrition in the Fortune 500, and you can see how companies have a, a half life now, which perhaps is very different to 20 or 30 years ago. >> Right, right, exactly. Well, Mike and Andrew, thank you so much for coming on theCUBE. This was a really fascinating discussion. >> Thanks. >> Thank you, good to see you again. >> I'm Rebecca Knight, stay tuned for more of theCUBE's live coverage of the AWS Executive Summit. (techno music)

>> Live from Las Vegas, it's theCUBE, covering the AWS Accenture executive summit. Brought to you by Accenture. >> Welcome back everyone to theCUBE's live coverage of the AWS executive summit, I'm Rebecca Knight, your host, we're here at the Venetian in Las Vegas. We have two guests for this segment, we have Mike Moore, Senior Principal Accenture Research, and Chris Wegmann, Managing Director Accenture AWS business group, thank both you so much for coming on theCUBE, and you for returning to theCUBE. >> Thanks for having me back, it's good to be back. >> So Mike I want to start with you and talk about your recent research which is entitled Discover Where the Value's Hiding: How to Unlock the Value of Your Innovation Investments. I like it, 'cause it just makes me think that innovation's just hiding somewhere in the corner, maybe underneath this desk. So talk a little bit about why companies can't find the innovation, and how they're failing at this. We'll get to the rays of hope later, but talk a little bit about what the problem is as you see it. >> Well, it certainly seems to be hiding for a lot of companies. Based on the research that we did, we found that over the course of the past five years large income companies like our own, and then also start-ups, have spent a combined 3.2 trillion dollars on innovation, which is obviously a huge sum. But when we looked at the rate of return on that investment, incumbent companies, we found it was actually declining by 27% over the course of that five year period. So there's a clear disconnect there in terms of what companies are doing. >> So, why are companies, why are so many companies not good at this? >> Well, we asked 840 executives from around the world exactly that question. And what we found is that for the vast majority of them who have been increasing their investments in innovation, they weren't seeing a great return, and what we found is that many of them were focused on incremental innovation, just small tweaks and adjustments to their products and services. But that's really not enough to get new customers in this day and age. But there was some reason for optimism, around about 14% of our respondents said that actually they were translating their investments in innovation into accelerated growth, they were outperforming their peers in terms of profitability, in terms of their market growth and they actually expected to continue to do that over the course of the next four to five years. >> So, I want to talk about that 14%, that rarefied group, but I want to bring you into the conversation Chris and just talk a little bit about the relationship between Accenture and AWS, and how you approach innovation, and how you help clients think about their innovation and driving ingenuity and creativity in their businesses. >> So Accenture and AWS have been partners for over twelve years, even before the first AWS service hit the market, Accenture was starting to use it in our labs at that point, and looking at how we could leverage S3 to really innovate on, and we've carried that tradition on for a while now. A couple years ago we sat down and really looked at what our enterprise customers were struggling with as they moved to the Cloud, and at that point innovation wasn't quite the topic yet, it was really how do we use Cloud to get better returns on our investment, better TCOs, things like that, and now we've seen that turn, as AWS has created more and more capabilities and solutions and offerings, our customers are really wanting to figure out how they innovate. They go and ask AWS "How do you innovate?" And that's their number one, one of their biggest EBCs is how do they innovate. So, we looked at it and said, that's great, how do we take that to the next level? How do we fix these failures that're happening and what we've seen is most customers are in this stop and go innovation traffic, I like to call it. There's people that're whizzing by them, the 14% are whizzing by them in the fast lane, so the question is how do you get them out of that stop and go traffic, into that fast lane, and, There's no lack of ideas, they have tons of ideas on how they can innovate, how they can use drones, how they can use all this. The ideas are out there, but taking those and turning those into operationalized assets that're continuously working, continuously growing, continuously maturing is where they struggle. >> And the question, when companies would ask you how do you innovate, I mean it is this question, but as you're implying it sounds as though it's a very, you have to have some discipline around it, there has to be real processes around the innovation, it's not just throw a bunch of creative people in a room together. >> No, that's great, you can do the creative people and they come up with the great ideas and there's no lack of those, but then you've got to operationalize those and go through the disciplining to take those, pick which ones are going to drive value, invest in those, operationalize those, and take them from a proof of concept or a pilot or whatever you want to call it and actually turn it into something that gets used every day, and what we've been focusing on with AWS is how do you get out of that, take what's out of that ideation stage, operationalize it using the full set of AWS services, and then how do you continue to run that and prove it going forward. >> So Mike, the 14%, what are they doing? What makes them different? >> Well I think there's lots of things that stand out, but there are really three things that came of the research, so firstly that group of companies is outcome led, as you were just saying, they're not just relying on the method, the genius in the garage tinkering away, but they're putting a real set of processes around the innovation activities that they're pursuing. Then they're linking those activities to clear operational and financial metrics. And then secondly they're disruption minded, so unlike the other companies that aren't performing well, they're really focused on investing in disruptive technologies that have the potential to create entirely new markets. And then finally, they're change orientated, so they're not just using innovation to develop new products and services, but they're also using it to drive more fundamental change across their organizations. And one of the principle changes that they're making is that they're becoming what we call network powered, so they're not just relying on their own internal innovation but they're drawing on a wide ecosystem of partners, like AWS, to really supercharge the rate at which they innovate. >> So those are the characteristics, what are you seeing on their ground, can you give us some specific examples of how they're taking those characteristics and what they're actually doing? >> I think you see companies set up and grow these innovation pods, so what we see customers doing is expanding those beyond just one pod. So, not just focusing on one part of their organization to do this, bringing that into a central location, creating a hub of pods and capabilities using everything, AWS services, using DevOps, doing all the cool stuff that's out there, but operationalizing that and getting to that center of excellence where they're actually seeing it end to end and they're not just jumping from one problem to then next. And once that graduates out, they have an organization waiting to take that on and continuing that journey while the next set comes in. So it's this process, it's this ongoing kind of chain of different problems coming in, being solved, and graduating out the other end. >> Is this a technology issue or is it a culture issue? >> The technology is there, I don't see it as a technology issue, I see it as a cultural issue, a change issue, a organizational issue, a resource issue, you got to find the talent that does that, you got to have the operational discipline to lead this stuff, and you have to go through that change. And we're seeing I think a lot of our customers struggle with that, and they want to learn how Accenture's done it, they want to learn how AWS has done it because obviously they've been very successful at it. >> And in terms of the cultural, the change management challenges that you're talking about, those are harder to overcome. So, do you have any best practices from your own experiences with it? >> We've obviously, Accenture's been in this game for a long time, whether it's innovation or whether we called it solution integration, whatever we called it, change was always a big part of that. So, a lot of those same change principles that we've used for twenty, thirty years still apply here. We see, you need very top down ownership and sponsorship, so from the very top down, whether that's the CEO, the CDO, the CIO, whoever it is they have to be 100% behind this, and have to be the cheerleaders. They have to be the people that're going to go get on stage, at re:Invent or other conferences and be that, this is how we're going, so you need that lead, and then you need very strong leadership underneath it that have gone through the journey before, this isn't the first time they've done it, they know where the potholes are in that road, they know what the signs are when they're going down the wrong way and how to get out of that. So you got to have those two key levels of experience. >> And to bring the others on-board. >> Absolutely, and they have to be the visionaries, they have to be the people guiding them through that, and you know, if you've got those people, if they're very strong-willed, very luminaries, those people will follow, and they'll follow them through that journey. And then they also got to go sell that to the rest of the organization, 'cause it's a change for the rest of the organization, the business is now much more engaged in that process, they're not just sending the requirements over the fence, they're very much engaged, they've got to understand and go through that agile transformation and understand when they're getting capabilities, what those capabilities are, so they need to go through that new operational paradigm that we're running in. >> So, finally, we're talking about innovation and then in particular AWS and Accenture, almost as the use cases here, how would you describe the innovation engine at AWS, Accenture in terms of the report that you've just published? It's obviously, I mean AWS is the biggest part of Amazon, I mean the high-growth engine of Amazon, and obviously a huge growth engine for Accenture too, so how would you characterize it Mike? >> Well, I think if we look at the three factors I was talking about before, being outcome led, being disruption minded, and being change orientated, then the relationship between Accenture and AWS really exhibits all of those three things, so in terms of being outcome led, Accenture has always been an organization that's laser focused on delivering results, delivering high performance, delivering value for our clients, and so is AWS. In terms of being disruption minded, we're innovating on the Cloud, using AWS to bring genuinely new and groundbreaking products and services to our clients. One of my favorite ones of those that we worked on in the UK, is our partnership with AWS and Age UK, which is a charity that helps the elderly and we're developing products and services for the elderly that helps them feel more connected to their family, and that's really opening up a brand new market. And then finally in terms of being change orientated, well, it's a relationship that really personifies being network powered and bringing the power there to multiple organizations that we can develop great products and services for our clients. >> Great, well Mike, Chris, thank you both so much for coming on theCUBE, it was a really fun conversation. >> Thanks for having us. >> Thanks. >> I'm Rebecca Knight, we will have more of theCUBE's live coverage of the AWS executive summit coming up in just a little bit.

you live from Orlando Florida it's the cube coverage conf 18 got to you by Splunk welcome back to comp 2018 this is the cube the leader in live tech coverage my name is Dave Volante I'm here with my co-host Stu minimun we're gonna start the day we're going to talk to some customers we love that John Morris here is the MIS program director at UConn the Huskies welcome to the cube good to see you and he's joined by Ryan O'Connor who's the senior advisory engineer at Splunk he's got the cool hat on gents welcome to the cube great to have you thank you thank you for having us so kind of a cool setting this morning is the Stu's first conf and I said you know when you see this it's kind of crazy we're all shaking our phones we had the horse race this morning we won so that was kind of orange yeah team are and team orange as well that's great you're on Team Orange so we're in the media section and the median guys were like sitting on their hands but Stu and I were getting into it good job nice and easy so Jon let's start with you start always left to start with the customer perspective maybe you describe your role and we'll get into it sure so as you mentioned I'm the director of our undergrad program Mis management information systems business technology we're in the school of business under the operations and information management department the acronym OPI M okay cool and gesture Ryan tell us about your role explain the Hat absolutely yeah so I'm a member of an honorary member of the Splunk trust now I recently joined Splunk about a month ago back in August and yeah and outside of my full-time job working at Splunk I'm also an adjunct professor at the University of Connecticut and so I helped John in teaching and you know that's that's kind of my role and where our worlds sort of meet so John we were to when I were talking about the sort of evolution of Splunk the company that was just you know okay log file analysis kind of on-prem perpetual license model and it's really evolved and its permanent permeating throughout you know many organizations but maybe you could take us through sort of the early days and it was UConn for a while what what was life like before Splunk what prompted you to start playing around with Splunk and where have you taken it what's your journey look like so about three years ago we started looking at it through kind of an educational lens started to think of how could we tie it into the curriculum we started talking to a lot of the recruiters and companies that many of our students go into saying what skillsets are you looking for and Splunk was definitely one of those so academia takes a while to change the curriculum make that pendulum swing so it was how can we get this into students hands as quickly as possible and also make it applicable so we developed this initiative in our department called OPI M innovate which was all based around bringing emerging technology skills to students outside of the general curriculum we built an innovation space a research lab and really focused in bringing students in classes and incorporating it that way we started kind of slowly different parts of some early classes about three years ago different data analytics predictive analytics courses and then that really built into we did a few workshops with our innovate initiative which Ryan taught and then from there it kind of exploded we started doing projects and our latest one was with the Splunk mobile team okay you guys had some hard news around now well today right yeah maybe take us through that absolutely wanted sure yeah I'll take that so we we teach a course on IOT industrial IOT at the University of Connecticut and so we heard about the mobile projects and you know the basically they were doing a beta of the mobile and application so we we partnered with them this summer and they came in you know we have a Splunk Enterprise license through Splunk for good so we're able to actually ingest Splunk data and so as part of that course we can ingest IOT data and use Splunk mobile to visualize it all right right right maybe you could explain to our audience that might not know spun for good absolutely yeah so spun for good is a great initiative they offer a Splunk pledge license they call it to higher education institutions and research initiatives so we're able to have a 10 gig license for free that we can you know run our own Splunk enterprise we can have students actually get hands-on experience with it and in addition to that they also get free training so they can take Splunk fundamentals one and two and actually come out of school with hands-on experience and certifications when they go into the job market that's John name you know we talk so much about them the important role of data and you know that the tools change a lot you know when we talk about kind of the next generation of jobs you're right at that intersection maybe you can give you know what what are what are the students what are they looking for what are the people that are looking for them hoping that they come out of school with you know yeah it's it's um you have two different types of students I would say those that know what they're looking for and those that don't that I really have the curiosity they want to learn and so we we try to build this initiative around both those that maybe they're afraid of the technology and the skills so how do we bring them in how do we make a very immersive environment kind of have that aha moment quickly so we have a series of services around that we have what's called tech kits the students come in they're able to do something applicable right away and it sparks an interest and then we also kind of developed another path for those that were more interested in doing projects or they had that higher level skill set but we also wanted to cultivate an environment where they could learn more so a lot of it is being able to scaffold the learning environment based off of the different student coming in so it's interesting my son's a junior in college at GW and he's very excited he's playing around with date he says I'm learning are I'm learning Tablo I'm like great what about Splunk and he said what's that yes so yeah then though it's a little off-center from some of the more traditional visualization tools for example so it's it's interesting and impressive that you guys sort of identified that need and actually brought it to two students how did that all how what was in an epiphany or was that demand from the students how'd that come about it was a combination of a lot of things you know we were lucky Ryan and I have known each other for a long time as the director of the program trying to figure out what classes we should bring in how to build out the curriculum and we have our core classes but then we have the liberty to build out special topics things that we think are irrelevant up-and-coming we can try it out once if it's good maybe teach it a few more times maybe it becomes a permanent class and that's kind of where we were able to pull Ryan in and he had been doing consulting for Splunk for a number of years I said I think you know this is our important skill set is it something that you could help bring to the students sure yeah yeah I mean one of the big courses we looked at was a data analytics course and we were already teaching with a separate piece of software not gonna name names but essentially I looked at it one for one like what key benefits does this piece of software have you know what are the students trying to get out of it and then just compared to one for one to Splunk like could Splunk actually give them the same learning components and all that and it could and and with this one for swum for a good license and all that stuff we could give them the hands-on experience and augment our teaching with that free training so and they come out of school they have something tangible they can say you know I have this and so that would kind of snowball once that course worked then we could integrate it into multiple other courses so you were able to essentially replicate the value to the students of the legacy software and but also have a modern platform exactly exactly yep yeah you know that and that was a what was like a Doug was talking about making jokes about MDM and codifying business processes and yeah it's a little bit more of an antiquated piece of software essentially you know and I mean it was nice it did a great job but there wasn't when we were talking to recruiters and stuff it wasn't a piece of software that recruiters were actually looking at so we said we were hearing Splunk over and over again so why not just bring it into the classroom and give them that so in the keynote this morning started to give a vision I believe they call it Splunk next and mobile things like augmented reality are fitting in you know how do you look at things like this what what how's the mobile going to impact you especially I would think yeah so when we kind of came up with our initiative we identified five tracks that both skill sets we believe the students needed and that and companies were kind of looking for a lot of that was our students would go into internships and say hey you know the the set skills that were learning you know they're asking us to do all this other work in AWS and drones and VR so as again it's part of this it was identifying let's start small five tracks so we started with 3d printing virtual reality microcontrollers IOT and then analytics kind of tying that all together so we had already been building an environment to try and incorporate that and when we kind of started working with the spunk mobile team there's all these different components we wanted to not only tie into the class but tied into the larger initiative so the goal of the class is not to just get these students the skills interesting interested in it but to spread that awareness the Augmented part is just kind of another feature was the next piece that we're looking in to build activities and it just had this great synergy of coming in at the right time saying hey look at this sensor that we built and now you can look at data in an AR it's a really powerful thing to most people so yeah they showed that screenshot of AR during the keynote and one of the challenges that we have at the farm so we're teaching that this is the latest course that we're talking about on an industrial IOT one of the challenges we have at that farm is we don't have a desk we don't have a laptop but we do have a phone in our pocket and we have we can put a QR code or NFC tag anywhere inside that facility so we can actually have we have students go around and you know they can put an iPhone upto a sensor or scan a QR code and see actual live real-time data of what those sensors are doing which is it's an invaluable tool inside the classroom and in an environment like that for sure so it's interesting able to do things we never would have been able to do before I want to ask you about come back to mobile yeah as you you just saying it was a something that people have wanted for a long time it took a while yeah presumably it's not trivial to take all this data and present it in a format and mobile that's simple number one and number two is a lot of spunky users are you know they're at the command center right and they're on the grid yep so maybe that worked to your advantage a little bit and that you know you look at how quickly mobile apps become obsolete hmm so is that why it took so long because it was so complicated and you had a user profile that was largely stationary yeah and how is that change yes honestly I'm not sure in the full history of the mobile app I know there previously was a new mobile app and I are there was an old mobile app and this new one though is you use it the new one yes oh so when we're talking about augmented reality that might be we may not been clear on that augmented reality is actually part of its features and then in addition we have the Apple TV app is in our lab we have a dashboard displayed on a monitor so not only can we teach this class and have students setting up sensors and all this but we can live display it for everyone to come in and look at all the time and we know that it's a secure connection to our back-end people walk into the lab and the first thing I see is this live dashboard Splunk data from the Apple TV based off of project we've been working on what's that well that's a live feed from a farm five miles off campus giving us all these data points and it's just a talking point people are like wow how did you do that and you know it kind of goes from there yeah and the farm managers are actively looking at it too so they can see when the doors are open and closed to the facility you know the temperature gets too high all these metrics are actually used by the you know that was the important part to actually solve a business problem for them you know we we built a proof of concept for the class so the students could see it then their students are kind of replicating another final project in the class class is still ongoing but where they have to build out a sensor for for plants to so it's kind of the same type of sensor kit but it's they're more stationary plant systems and then they have to figure out how to take that data put it into Splunk and make sense of it so there's all these different components and you get for the students get the glam factor you can put it in a fishbowl have the Apple TV up there exactly and that's I mean part of it when we when we started to think about in ishutin you know it was recruitment you know how do we get students beyond that fear of technology especially kind of coming into a business school but it really went well beyond that we aligned it with the launch of our analytics minor which was open to anyone so now we're getting students from outside the school a bit liberal arts students creating very diverse teams and even in the class itself we have engineers business psychology student history student that are all looking to understand data and platforms to be able to make decisions so there's essentially one Splunk class today instead of a Splunk 101 there this semester there's there's a couple classes that are actually using Splunk inside the classroom and I mean depends on the semester how many we have going on that are actually there's three the semester I sorry I misspoke there we have a another professor as well who's also utilizing it so so yeah we have three three classes that are essentially relying on Splunk to teach different components or you know is it helped us understand is it part of almost exclusively part of the analytics you know curriculum or is it sort of permeate into other Mis and computer science or right now it's within our kind of Mis purview trying to you know build all their partners within the university and the classes they're not it's not solely on spunk spunk is a component of you the tool so it's like for example the particular industrial IOT course it is understanding microcontrollers understanding aquaponics and sustainability understanding how to look at data clean data and then using Splunk as a tool to help bring that all together yeah it's kind of the backbone you know love it and then and I mean in addition to I just wanted to mention that we've had students already go out into the field which is great and come back and tell us hey we went out to a job and we mentioned that we knew Splunk and we were you know a shoo-in for certain things once it goes up on their LinkedIn profile and start getting yeah I mean I again I would think it's right up there with I mean even even more so I mean everybody says and right and our day it was SPSS now it's our yep tableau obviously for the VIS everybody's kind of playing around but spunk is a very you know specific capability that not everybody has except every IT department on the planet exactly coming out of school you take a little bit deeper you either you find you find that out yeah cool well great work guys really thank you guys coming on the cube it was great to meet you I appreciate it incoming all right you're welcome all right keep it right - everybody stew and I will be right back after this this is day one of cough 18 from Splunk this is the cube [Music]

(upbeat music) >> Announcer: Live from Madrid, Spain, it's theCUBE. Covering HPE Discover Madrid 2017. Brought to you by Hewlett Packard Enterprise. >> Hi everybody, welcome to Madrid, Spain. My name is Dave Vellante, and this is theCUBE, the leader in live tech coverage. We're here, this is day one of HPE Discover Madrid, the European version of the event that we cover in the summer, in the spring, in Las Vegas. I'm here with my cohost, Peter Burris, and Bob Moore is here, he's the director of server software and product security at HPE, and he's joined by good friend Patrick Osborne, who runs product marketing and management for the storage group at HPE. Gents, welcome to theCUBE. >> Good to be here, Dave, Peter. >> Yeah, very happy to be here. >> Dave: Always good to see you Did you bring your sax? >> Not this time, my friend. (laughing) >> We had a lot of fun. Where were we in New Orleans last year? >> Oh yeah, it was great. >> And you're an awesome sax player, we love it, big fan, and you're a bass player, we got more sax, more horns over there. So, I digress. >> Patrick: You need a CUBE band (laughing) >> We need a CUBE band. >> Bob, we talked this spring in Las Vegas, you guys made a big deal about the silicon-level security, you made some innovations there. Give us the update on why, again, that's so important, and how that's been received by customers. >> Yeah, well I think, answer the second part of the question first, it's really resonating pretty well with customers. Honestly, as we get to them, and we describe the level of cryptography we have, down right into the hardware, the firmware, down into our silicon, those customers that are concerned with security, and frankly, all customers are now, really does resonate with them pretty well. And the reason that it's important is because tying all of that security down into a bedrock foundation provides that ability to then leverage in or pull in other objects like storage and provide that security without any increase in latency but also the access and the shared access, being able to do that across multiple platforms, do it securely, and have that sharing capability like we all need to have to keep our IT infrastructure running. So it's really critically important, still, to this day, HPE is the only server manufacturer that's able to do that down into the silicon level that we're talking about here. So we're quite proud about that. And it's allowed us to claim the world's most secure industry standard servers and now, of course, today we're branching out with other technologies across our storage platform and including those into our security strategy. >> So, how does it, Patrick, relate to what you guys are doing on the storage side? >> Yeah, so I think it's a really good complementary solution and the fact that we can provide the silicon root of trust on the infrastructure level, and then on the storage side, we provide some similar capabilities at the infrastructure level, with encryption and other techniques that we have, and then, we assist customers in being able to, in a number of different cases, being able to take, for example, snapshots in backup, move those offsite, or even into the cloud, encrypt those, so you have essentially a silicon-rooted trust on the infrastructure side for your operating system and your firmware. And then you have essentially a golden image at a point in time of your data, which is a pretty valuable asset. So combine those two, we're able to help customers with a pretty aggressive RTO, and RPO, to be able to recover, if they'd been breached, or when they get breached, essentially. So we have some great examples here today in the show, of some customers that have used combinations of things like, the Gen10 servers, 3PAR, and StoreOnce, to achieve that level of recovery, in, not days, in, basically in hours, or even faster. And then we have some other technologies where you can set up a media break, essentially send all that data out to the cloud, and completely have a self-contained, encrypted copy of your data to recover from. So we're providing a number of different solutions, all the way up and down the stack for customers to be able to help to recover very quickly. >> So obviously security's been in the news lately, the huge Equifax breach, you go back to the spring, WannaCry, and ransomware, >> Patrick: Yep. >> So let's talk about ransomware specifically. How do you guys help a customer sort of address that. What's the, there's no silver bullet. You hear talk about air gaps, you guys are talking about >> Patrick: Right. >> silicon-level security, What's the prescription for customers? >> Well, I'm glad you asked that, because ransomware really is on every customer's mind these days. And it is, because it's gone up, ransomware is so lucrative and profitable, it's gone up by 15 full, 15 times in the last two years, to the point where it's cost companies five billion dollars in 2017, and by 2019, a company will be infected by ransomware every 14 seconds, so it's just really huge. And not only, and we don't encourage paying the ransom, but the ransom, if you paid it, would be expensive, but the downtime that you experience in recovering can be really expensive for companies as well. So this ability to recover from ransomware, or ransomware neutralizer, which is what we're talking about and announcing here today, is really new and a revolutionary way to recover in a systematic, orderly fashion, starting with the firmware that we talked about, that's anchored down in the silicon, so we recover that firmware, in case that ransomware malware virus has migrated. Because the hackers are getting so incredibly ingenious these days, that that malware can hide inside the firmware and will go everywhere, the tentacles will go everywhere, but we start the recovery with a firmware so you've got that firm foundation, routing out any remnants of the malware. And then on top of that, new today, we're announcing the fact that we can then recover the server settings that take days, sometimes weeks to set up initially, and that'll be recovered and restored automatically. Then we restore the operating system through an ISO site, along with the applications and then finally, we bring the data back, as Patrick was mentioning, we do that relatively quickly. We're demonstrating that here this week at Discover Madrid. And it really does allow customers to avoid having to pay the ransom, we want them to be able to recover, do it quickly and easily, without paying the ransom, and that's what we help. >> But you mention the word "trust," which is one of the most increasingly important words in the tech industry. We're in Madrid, GDPR is going to start moving in into a force in the first quarter of next year. >> Bob: May 2018. >> So, second quarter. And it's going to create some fair amount of attention, not just here in Europe, but on a global basis. I was talking to an expert who suggested that if the Equifax breach had occurred in Europe, under GDPR, it would not have been just embarrassment, it would have been about 60, 70 billion dollars worth of funds. >> Bob: Right. >> So we're talking not just about nice things to have, we're talking about, over the course of the next five years, you have to have this level of capability inside your infrastructure, or you will be out of business. >> I think it's true, absolutely. The GDPR, the penalties associated are so severe with that, up to 20 million dollars, or four percent of the annual revenue of the parent company, so it can just be massively impactful, financially impactful, hurtful to the companies. We're talking today, and this week, about GDPR, and how we help companies get ready for that, and you mention the Equifax breach, actually, we have, with our HP Gen9 and Gen10 solutions server networking and storage, applied the NIST 800-53 controls to that, and if they had applied those and used our solution, we believe that, after having looked at the Equifax breach, that would not have happened, had they followed the security controls that are in NIST. There's a lot of articles published about how NIST can help companies get ready for the GDPR in Europe, and so we've got the NIST controls, we went through all the time, energy, and funding to create the NIST security controls that will help a hundred percent of those applied to the ISO certification, ISO 27000-1, 27000-2, which then lends itself to being GDPR compliant. So, not only do we help customers through this great new technology that we have in the silicon-rooted trust, and that's helpful in getting ready for the GDPR, but also these NIST controls. >> But it's also that it's also that the well the conversations that we're having with CIOs is that GDPR, even though it's centered here in Europe, is likely to have an effect on global behavior. And so, one of the things that they're looking for is, they're looking for greater commonality in the base infrastructure about how it handles security, so that they can have greater commonality in how their people do things, so they can be better at targeting where the problem is, when the problem happens, and how to remediate the problem. Talk a little bit about how more commonality in the infrastructure, especially when you talk about storage, which is increasingly the value proposition, is how you share data is going to liberate resources elsewhere in the business to do new and better things faster. >> I think for, from the HPE perspective, you're not going to solve GDPR with any specific point product. Right? And that's not, it's not really our message to the market, that, you implement this and you're going to go satisfy those requirements. It's definitely part of a solution, but what we've been trying to do is, you see, we've got the silicon root of trust on the server side, and a number of security features, and we're talking about how we integrate that with the storage. We're starting to bring together more of a vertically oriented stack, that includes all those pieces and they work together. So instead of having a security or commonality layer at the server layer, at the networking layer, at the storage layer, thinking about it as a service that's more vertically oriented through the stack, where you're able to take a look at all aspects of the networking, what's going on with the firmware and the operating system and all the way down to essentially your secure and most important data. >> Peter: Securing the data >> Exactly. >> And not the device. >> Exactly. Exactly. And so for us, you see it in themes for for 3PAR, for SimpliVity on the hyperconverged area, and all the converged systems on the compute side, we're really providing integrated security and integrated data protection that is inherently secure with encyryption and a host of other techniques. So really, we're trying to provide it from the application level on down through the infrastructure, a set of capabilities within the products that work together to provide a little bit more of a secure infrastructure. >> One of the things we talked to Bill Philbin about on theCUBE recently was, and Patrick, I'm sure you've heard this, maybe you too as well, Bob, but boo-boos happen now, today, really fast. So they replicate very quickly. So how do you deal with fast boo-boo replication and sort of rolling back to the point where you can trust that data? >> There's a couple techniques and innovations that we brought within the storage realm, in terms of integrating that whole experience, so our big thing is, on the storage side, has been how can you provide an experience from all-flash on-prem out to the cloud, from a data perspective, and have all that integrated so we've got a number of things that we've actually announced here at Discover, in terms of 3PAR, all-flash, and Nimble, being able to federate that primary storage, with your secondary storage, on-prem, and then being able to have that experience go off-prem, into the cloud, so you do have a media break and a number of things. I think, from a solution perspective, integrating with some of our top-tier partners on the availability side, like Deem, for example, it gives you that really holistic application-level view, in the context of virtualization, it's something that helps do the very rich cataloging experience, and pieces. >> So I wonder if we could talk about a topic that's been discussed in our communities, which is the biggest threat within cyber is the weaponization of social media. You've sort of seen it with fake news, and Facebook, and I wonder if you guys are having similar conversations with customers and even ransomware. You look at WannaCry, it was sort of state-sponsored, and actually not a lot of money went back >> Patrick: Right. >> To the perpetrators, maybe it was a distraction to get other credentials. And you're seeing different signatures of Russians, very sophisticated hackers, they target pawns and make 'em feel like kings, and then grab their credentials, and then go in and get critical data. So when you think about things like the weaponization of social media, how can you guys help, sort of, detect what's going on, anomalous behavior, and address that? You've got silicon level >> Right. >> You've got the storage component. Do analytics come into play? Is there a whole house picture that you can help customers >> Yeah, I think that's the next level. It's almost an iterative process as soon as we've developed a protection, or the ability to detect a cybersecurity breach, is then the hackers try to outdo that, and so we're continually leapfrogging, and I think the next step is probably with machine learning. We're starting to actually deploy some of that at HPE, that artificial intelligence, and we have some of that now with our storage, our Nimble storage, as well as our Aruba Networking with the technologies that Aruba has with IntroSpect, can now look at the communication inside of a network and determine if there's nefarious behavior, and watch the behavior analytics, as well as the signatures that are going on inside the network, and actually, then communicates with ClearPass, and can proactively take some charge of that and rule out that user that's potentially a bad actor before any damage is really done. Same way on, with the storage side, >> Patrick: Yep. >> With the InfoSight that has great, in fact, so great of AI intelligence, that we're actually sharing as we look at ransomware viruses, they're looking at the signatures that those leave, and the trails that ransomware leaves behind, so that the storage systems can actually proactively route that out with machine learning and artificial intelligence. That's where we're headed with HPE. >> But it's, it's not only, it's not only finding ways to fix the boo-boos, it's acknowledging or recognizing that the boo-boos occurred. So how is this new capability facilitating, or increasing the speed with which problems are recognized? >> I think one of the important points that Bob made is that we are, we're announcing this week, on the storage side, some concepts around AI for the data center, and specifically, around our predictive analytics with InfoSight, and applying that from Nimble to the 3PAR systems, and then setting out a vision that is going to basically enable us to use that AI at the infrastructure layer, across other areas within the portfolio. Servers, networking, and for, at the speed at which this is moving, you can't solve this at the human level, right? So for us, to be able to whitelist and blacklist customers, based on our learning across a very large install base, if you think about the amount of compute nodes and the amount of storage that we sell as a infrastructure company, you can learn and be enabled to proactively help customers avoid those situations, that's something we're actually implementing today. >> And let me follow up with that, because it's a great lead-in or tie-back to GDPR that we were discussing. >> Yep. >> Because there's reporting requirements within 72 hours, right, >> Yep. >> That GDPR says that you've got to report that you had a breach, and how do you report that if you're not certain? Well, with our silicon-rooted trust and the Gen10 servers, we actually are monitoring all that server essential firmware every 24 hours. Now some of our competitors monitor, or check the firmware, one time when you boot up the server, and never again until you, maybe reboot the server, right? But we're doing, at HPE, that check every 24 hours, and that's an automated process. And so, you ask, how can be detected? Well, we can detect that, because you'll get an alert, coming back to the user of the server, that there's been a breach, and that can be reported. >> We got to go. I'm glad you mentioned automation, because that's a big factor, >> Bob: Yeah. >> Using false positives, because people just don't have time, they're drinking from the fire hose. Bob, Patrick, thanks very much for coming to theCUBE. >> Great, thanks so much for having us. >> Dave: Enjoy the week. >> Thank you so much, we appreciate it. >> All right, keep it right there everybody, we'll be back with our next guest. This is theCUBE. We're live, from HPE Discover in Madrid. We'll be right back. (upbeat music)

>> Male Voice: From the Fairmont Hotel in the heart of Silicon Valley, it's the Cube covering when IoT Met AI: the Intelligence of Things brought to you by Western Digital. >> Hey welcome back here everybody. Jeff Frick with the Cube. We're in downtown San Jose at the Fairmont Hotel at a small event talking about data and really in IoT and the intersection of all those things and we're excited to have a little startup boutique here and one of the startups is great enough to take the time to sit down with us. This is Shaun Moore, he's the founder and CEO of the recently renamed Trueface.ai. Shaun, welcome. >> Thank you for having me. >> So you've got a really cool company, Trueface,ai. I looked at the site. You have facial recognition software so that's cool but what I think is really more interesting is you're really doing facial recognition as a service. >> Shaun: Yes. >> And you a have a freemium model so I can go in and connect to your API and basically integrate your facial recognition software into whatever application that I built. >> Right so we were thinking about what we wanted to do in terms of pricing structure. We wanted to focus on the developer community so we wanted tinkers, people that just want to play with technology to help us improve it and then go after the kind of bigger clients and so we'll be hosting hack-a-thons. We just actually had one this past week in San Francisco. We had great feedback. We're really trying to get a base of you know, almost outsource engineers to help us improve this technology and so we have to offer it to them for free so we can see what they build from there. >> Right but you don't have an opensource component yet so you haven't gone that route? >> Not quite yet, no. >> Okay. >> We're thinking about that though. >> Okay, and still really young company, angel-funded, haven't taken it the institutional route yet. >> Right, yeah, we've been around since 2013, end of 2013, early 2014, and we were building smart home hardware so we had built the technology around originally to be a smart doorbell that used facial recognition to customize the smart home. From the the trajectory went, we realized our clients were using it more for security purposes and access-control, not necessarily personalization. We made a quick pivot to a quick access control company and continue to learn about how people are using facial recognition in practice. Could it be a commercial technology that people are comfortable with? And throughout that thought process and going through and testing a bunch of other facial recognition technologies, we realized we could actually build our own platform and reach a larger audience with it and essentially be the core technology of a lot cooler and more innovative products. >> Right, and not get into the hardware business of doorbells >> Yeah, the hardware business is tough. >> That's a tough one. >> We were going to through manufacturing one and I'm glad we don't have to do that again. >> So what are some of the cool ways that people are using facial recognition that maybe we would never have thought about? >> Sure, so for face matching - The API is four components. It's face matching, face detection, face identification, and what we call spoof detection. Face matching is what it sounds like: one-to-one matching. Face detection is just detecting that someone is in the frame. The face identification is your one to act so your going into a database of people. And your spoof detection is if someone holds up a picture of me or of you and tries to get it, we'll identify that as an attack attempt and that's kind of where we differentiate our technology from most is not a lot of technology out there can do that piece and so we've packaged that all up into essentially the API for all these developers to use and some of the different ideas that people have come up with for us have been for banking logins, so for ATMs, you walk up to an ATM, you put your card in and set up a PIN so to prevent against fraud it actually scans your face and does a one-to-one match. For ship industries, so for things like cruise ships, when people get off and then come back on, instead of having them show ID, they use quick facial recognition scans. So we're seeing a lot of different ideas. One of the more funny ones is based off a company out in LA that is doing probation monitoring for drunk drivers and so we've built technology that's drunk or not drunk. >> Drunk or not drunk? >> Right so we can actually measure based on historical data if your face appears to be drunk and so you know, the possibilities are truly endless. And that's why I said we went after the development community first because >> Right right >> They're coming to use with these creative ideas. >> So it's interesting with this drunk or not drunk, of course, not to make fun of drunk driving, it's not a funny subject but obviously you've got an algorithm that determines anchor points on the eyes and the nose and certain biometric features but drunk, you're looking for much softer, more subtle clues, I would imagine because the fundamental structure of your face hasn't changed. >> Right so it's a lot of training data, so it's a lot of training data. >> Well a lot of training data, yeah. We don't want to go down that path. >> So a lot of research on our team's part. >> Well then the other thing too is the picture, is the fraud attempt. You must be looking around and shadowing and really more 3D-types of things to look over something as simple as holding up a 2D picture. >> Right so a lot of the technology that's tried to do it, that's tried to prevent against picture attacks has done so with extra hardware or extra sensors. We're actually all cloud-based right now so it isn't our software and that is what is special to us is that picture attack detection but we've a got a very very intelligent way to do it. Everything is powered by deep learning so we're constantly understanding the surroundings, the context, and making an analysis on that. >> So I'm curious from the data side, obviously you're pulling in kind of your anchor data and then for doing comparisons but then are you constantly updating that data? I mean, what's kind of your data flow look like in terms of your algorithms, are you constantly training them and adjusting those algorithms? How does that work kind of based on real time data versus your historical data? >> So we have to continue to innovate and that is how we do it, is we continue to train every single time someone shows up we train their profile once more and so if you decide to grow a beard, you're not going to grow a beard in one day, right? It's going to take you a week, two weeks. We're learning throughout those two weeks and so it's just a way for use to continue to get more data for us but also to ensure that we are identifying you properly. >> Right, do you use any external databases that you pull in as some type of you know, adding more detail or you know, kind of, other public sources or it's all your own? >> It's all our own. >> Okay and I'm curious too on the kind of opening up to the developer community, how has that kind of shaped your product roadmap and your product development? >> It - we've got to be very very conscious of not getting sidetracked because we get to hear cool ideas about what we could do but we've got our core focus of building this API for more people to use. So you know, we continue to reach out them and ask for help and you know if they find flaw or they find something cool that we want to continue to improve, we'll keep working on that so I think it's more of a - we're finding the developer community likes to really tinker and to play and because they're doing it out of passion, it helps us drive our product. >> Right right. Okay, so priorities for the rest of the year? What's at the top of the list? >> We'll be doing a bigger rollout with a couple of partners later on this year and those will be kind of our flagship partners. But again, like I said, we want to continue to support those development communities so we'll be hosting a lot of hack-a-thons and just really pushing the name out there. So we launched our product yesterday and that helped generate some awareness but we're going to have to continue to have to get the brand out there as it's now one day old. >> Right right, well good. Well it was Chui before and it's Trueface.ai so we look forward to keeping an eye on progress and congratulations on where you've gotten to date. >> Thank you very much. I appreciate that. >> Absolutely. Alrighty, Shaun Moore, it's Trueface.ai. Look at the cameras, smile, it will know it's you. You're watching Jeff Frick down at the Cube in downtown San Jose at the When IoT Met AI: The Intelligence of Things. Thanks for watching. We'll be right back after this short break.

>> Announcer: Live from Las Vegas it's theCUBE. Covering HPE Discover 2017 Brought to you by Hewlett-Packard Enterprise. >> Okay welcome back everyone, we're here live in Las Vegas it's theCUBE's exclusive coverage of HPE Discover 2017. HP Enterprises premier show, it's theCube on our third day I'm John Furrier, my co-host Dave Vallante. And our next guest Bob Moore returning back, Director of Server Software Private Security, he's got the hottest product, he's here on the show. We're going to go do a deeper dive. And Jason Shropshire SVP, CTO of InfusionPoints. Welcome back welcome to theCUBE. >> John, thank you Dave. You're the talk of the town here on the show with the simple messaging that is clean and tight. But outside of that, from a product stand point is really some of the security stuff you guys are doing in the Silicon. >> Bob: It is. >> In the server with Gen 10, pretty game changing, we've been curious, we want more information. >> Bob: Yeah. >> John: Give us some more update, what's the update? >> Glad to do that we're really proud of the announcement of course it's a big bold announcement this week. Claiming ourselves the world's most secure industry standard server. So that's big, that's huge, that's based on this new revolutionary security technology that we've been developing frankly over the past couple of years. So it's been two or three years in the making. A lot of hard work, we actually started to look at what type of security trends were happening, and what we might have to do to protect the servers. And we've come up with a game changing capability here. And it's one thing for us to say it internally at HPE, but we are so certain that we are in a great security position that we went external and found a security firm outside, that independently could look at it and do some compare, contrast testing with a competitive unit, so. >> So let's drill into that, actually I had some other questions on the industry in terms of what is going on at the chip level. Always on security is kind of a theme we've heard in the past from some of your competitors, but lets get into some of the competitive analysis. What do you guys see in the benchmarks. Jason, what do you guys discussing , because at the end of the day, claims are one thing. No offense to HP, you're kind of biased of course. We have folks on from the marketing team as well. Where's the proof in the pudding? >> Oh yeah, well one thing that we know for sure is that threat is real, right, with firmware. And it was great for us to analyze HP's new technology. We had on the bench two different beta units. >> John: You guys are the ones who did the benchmark. >> Jason: Yes the analysis. >> Independent. >> Independent yeah, FusionPoints is a cyber security firm independent from HP, they approached us to do the testing. >> John: Okay. >> We have head analysts that do this sort of thing all the time for our customers. >> So take us through what happened. >> Yeah so they procured for us three competitor servers. Sent them to our shop. We set them on a bench, all side by side. From what I can tell, no one's ever really done a test like that, you know, within the server industry. It was very exciting. There's been a lot of benchmarking done and performance, things like that. But from a black hat stand point, to actually look at the hardware, that hardware level testing, we couldn't find any examples of anyone doing it. I thought that alone was just evidence that HP was very serious about security and they knew what they had. So. >> John: You guys getting your answer, because you know the malware, and all the ransomware going on. People are going through elaborate lengths. >> Jason: Absolutely. >> Business model, organized teams, this is a really orchestrated security market now, with the black hat guys out there, really hacking away at every angle. >> Yeah well, you know we saw evidence that firmware issues and exploits are here to stay. The Vault7 release that happened recently showed us that there are exploit kits. Intel security released within a day a tool to let you do firmware validation. But to do that you have to take your server offline and build a gold image of what that firmware should look like. And then compare like a week later if you think you might have had a breech. You have to take your server down and compare it against that gold image. And who has the time to do that? But what we found in analyzing the Gen 10 server is HP has built this in, where this can be done in real time, while the server is running. No performance hit, no down time. It really is a revolutionary game changer I think for firmware security. >> So Bob, can you explain what IP you developed in Silicon that Intel, where do they leave off and you pick up? >> Sure, sure because Intel has some great security technology. And we actually support a lot of Intel technology. Their TXT, their Trusted Execution Technology as part of our Gen 10 servers. But what we've done at HPE is we've really taken it multiple steps further than that and we've developed. Because we're in a position where we develop our own custom HPE iLo-silicon chip, we're able to anchor what we actually do, imbed the cryptographic algorithms into that, and we anchor all the server's essential firmware. Right, think of it as anchoring it down into the bedrock. So there's really no way you can get in and breach that. And even if you did, instead of taking it offline like Jason was talking about here. We have the ability to not only provide that protection, but we would detect any type of malware or virus that gets in. And then frankly, we can recover that, almost immediately within a few minutes. In fact we're demonstrating that here during Discover this week. >> Is there anyplace online where people can get information, people watching, probably curious. >> Bob: Sure >> You can just give them the URL. >> Yeah just naturally it's our HPE.com/security. And that where there we've got some white papers there and other things there. >> So you say you can recover universally instantaneously. >> Bob: Yes. >> And you do that by what, fencing certain resources or... >> Yeah well what we've done, is we verify as the server is running, we're doing a runtime for more validation. So we're checking that firmware, making sure it's free of any malware, viruses, or compromised code. Completely perfect in original shape, like when we ship it from the factory. And we're storing in another location inside the server, a secure copy of that. Think of it as log box, inside the server, where it can't be found unless we need it to go into recovery mode. Then we draw from that, we've checked it daily, we've stored it there, we know it's authentic, and we can pull that back into recover in case something does happen to the server. >> And then asynchronously reclaim that wasted resource, clean it up and bring it back online. >> We can, we can recover the server, the firmware, toward the end of the year, we'll be recovering the operating system as well. Also we've got a really holistic way to get that recovered. When we talk to customers, a real big concern, and sometimes it's called bricking a server, you've got a bricked server, something that just won't operate. And it's important because 60% of small businesses that suffer a security breach, go out of business within six months, so it can be huge that lack of cashflow for customers. It's that denial of service, that disruption of business. Well we prevent all that, because we can not only protect the server, but then recover from a breach. >> So the anatomy of that breach, can we go through a common use case? So malware gets in, it gets into the server, it's hiding, typically you don't know about it. And in this new scenario with your Gen 10. You'll be able to identify that. >> Bob: That's right. >> To protect it, okay. And if I understand, the business impact of the problem you're solving is, not only are you sort of automating that protection, but you're also eliminating, a lot of wasted time, and downtime, and accelerating the response. >> Yeah I think that's what Jason was talking about earlier. Normally, if you're server gets infected, you completely take it off line and then do a manual recovery. And customers still have the choice to do that, but in our case we can recover immediately within a few minutes if something happens and gets a breach. >> Those types of exploits are typically in the data plane as well. With firmware you can't even really detect that you've been hacked. So down in the firmware virus scanners, those things don't work. So if you have a BIOS exploit, that is on either the iLO, that would be on the BMC the baseboard management controller. And undetectable by the operating system. >> That's crazy because it's a clean haven for hackers. I mean they know how to get in there, once you're in there, you're in. >> I don't know if a lot of customers realize this but the first thing when you turn a server on, there first thing that comes on is the firmware. In our case it's the iLO firmware. Over a million lines of the firmware code run before the operating system even starts. So that can be a cess pool for a trojan horse. And the research shows a virus, somewhat analogous to a human virus, it can stay there, hibernate in there for months, maybe even a year or more until it springs forth and opens up the passwords or bricks your servers, or does some nefarious thing. >> A cesspool from the customer standpoint, from a hacker is like going to the beach. Pina Coladas, you're clean you're down there having fun. >> Well what's your stats? The average time to detect an intrusion is over 200 days. >> Bob: That's right yeah. >> So essentially, you're detecting it instantaneously. >> We can, we run that runtime firmware validation on a regular basis, can be run as much as everyday, and so you'll know almost immediately. Which is really great because of a lot of regulatory bodies want to know if a breach has occurred. So this gives customers the ability to know somethings happened to them. >> Jason I want to challenge the claim here, because first of all I love the bravado. Yeah, we're bad ass, we're number one. >> We know that. >> What is the, how did the leaderboard come out? What was the results? Did HP come out number one? >> Oh absolutely. >> What's the lead, what's the gap, talk about the gap between HP and other servers. Did they send you the best servers? What was the benchmark, I'm sure you did your due diligence, take us to more of the results. >> Sure, sure, so yeah again we were comparing all the servers side by side. A test that had never been done from what I'd seen. When we looked at by feature, by feature, and started analyzing things. We sort of broke down and we saw we really had two different angles that we were looking at. The penetration test as aspect. What we were looking for vulnerabilities in the firmware, at the physical layer, at the network layer. They passed that with flying colors. We found a few minor issues that they jumped on and resolved for us in a matter of hours or days. And then the other aspect was a feature by feature comparison that we looked at. We looked at the silicon retruss obviously and we saw what the others were doing there. At best the other guys were using firmware to validate firmware. The obvious issue with that is if the firmware is compromised it's not trustworthy. >> Spoof, yeah, yeah. >> It's in no position to validate and verify. >> It's like Wallstreet policing itself. >> Jason: Yeah, can't trust that, They have a revolutionary intrusion detection switch on the Gen 10, that actually detects. If the lid is lifted on the server, anywhere from when it leaves the factory to the garage of the installation point, server doesn't have to be plugged in like the other guys. >> So if it's just a physical casing breach, >> Jason: Exactly. >> What happens there, flags the firmware, makes a note, does it shut it down? What happens? >> It makes a note, it puts it in the log entry so you can tell if that server has been tampered with in transit. >> So the insider threat potential should go away with that. >> Right, physical access, you don't have to worry about that because we can verify that server gets to the customer in it's unique, original, authentic condition. Because even though the power is off that is going to register and auto log an alert if that chassis has been opened. >> So I can't go to the vault of the Bellagio, like they did in Ocean's Eleven and put my little, and break into the server and you know go in there. >> Bob: Exactly. >> Okay, now back to the results. So the other guys, did they all pass or what. >> Well we did find some issues that we're looking at and doing some further testing on. >> So we're going to be polite and respect the confidentiality you have the ethos of security as you know sharing data is a huge deal, and it's for the integrity of the customer that you have to think about so props for that. For not digging into it, we'll wait for an official report if it does come out. Alright, so I got to ask you a personal question Jason. As someone who is in the front lines. You know every time there's a new kind of way, whether it's Bitcoin and Block chaining, you see a slew of underbelly hacking that goes mainstream that people are victimized. In this case firmware is now exposed, well known. >> Jason: Yeah. >> What as a professional, what gets you excited, and what gets you alarmed if anything about this? What new revelations have you walked away with from this? >> Well it's just how pervasive this issue is. You know the internet of things has exploded the number of IP devices that are out there. Most of them have, firmware issues, almost all of them have firmware issues. And we've just now seen bot nets being created by these devices. Cameras, IP cameras and things like that, that become attack platforms. So I just want, one of the things that impressed me very much about HP's approach here is that they're being a good corporate citizen by making a platform that's going to be implemented in tens of thousands IP addresses. Those systems I think will be much more secure. Again it can't become an attack platform for other people, for attackers to abuse. >> So the surface area, so your point about IOT. We always talk about the surface area of attack vectors. And that vector can then be minimized at the server level, because that's like the first mile in. >> Right we come and really refer to that as the attack vector or the attack surface. And so we narrow that attack surface way down. >> Can you even subjectively giVe us a sense as to how much of the problem this approach addresses? I mean is it 1%, 10%, 50% of the attacks that are out there? >> I think the important thing here is moving, shifting the bar. I've likened this, what HP is doing here to what Bill Gates did 15 years ago with the Microsoft memo. I mean that really revolutionized operating systems security within Microsoft and I think it had a ripple effect out into the industry as well. So I think HP is really pushing the bar in the same way but for firmware, instead of the operating system level that was the paradigm 15 years ago. >> And I think you'll find on our website we put some of the studies actually, and it's over half, I think it's 52% of the firms that responded have had a breach or malware virus in their firmware. So over half of those, and 17% had a catastrophic issue with that, it really is more pervasive. We've seen a lot of news about the data plane level, where thefts are taking place at the application level of the operati6ng system. And we've got to pay attention to the firmware layer now because that's like I said, a million lines of code in there running. And it could be an area where a trojan horse can sit, and we essentailly narrow that attack surface. We're also delivering with the Gen 10, the highest, the strongest set of security ciphers available in the world today. And that's a commercial national security algorithms. We're the only ones to support in our server, so we're proud of that. >> Well Bob and Jason thanks so much for sharing the insite. It's super exciting and relevant area, in the sense that it's super important for businesses and we're going to keep tracking this because the Wikibond team just put out new research around true private cloud, showing the on prim, cloudlike environments will be 260 billion dollar market. That's new research, that's groundbreaking, but points to the fact that the on pram server situation is going to be growing actually. >> Jason: For sure. >> So this is, and with cloud there's no perimeter so here you go, firmwares, potential exposure you solved that problem with good innovation. Thanks so much for sharing. >> Thanks you guys. >> Thank you. >> The inside Jason and Bob here on theCUBE talking about security servers, attack vectors, no perimeter, it's a bad world out there. Make sure you keep it protected of course. This is CUBE bringing you all the action here at HPE Discover. We'll be right back with more live coverage after this short break. I'm John Furrier, Dave Vellan6te. Be right back after this short break, stay with us.

>> Narrator: Live from Las Vegas, Nevada, it's theCUBE covering Accelerate 2017 brought to you by Fortinett. Now, here are your hosts, Lisa Martin and Peter Burris. >> Hi, welcome back to theCUBE. We are SiliconANGLE's flagship program where we go out to the events and extract the signal from the noise. Today, we are with Fortinet at their 2017 Accelerate event in Las Vegas. I'm your host, Lisa Martin, and I'm joined by my cohost, Peter Burris. We are fortunate right now to be joined by George Moore. George is the technology, excuse me, the CSO for Microsoft Azure who is a big technology alliance partner for Fortinet. George, welcome to theCUBE. >> Nice to have you, thank you. >> We are excited to have you on. You are, as you mentioned, the CSO at Azure, but you are the CSO for all of the Azure computer services. You are one of the founders of the Azure engineering team from back in 2006, and we were talking off-line. You hold over 40 patents in things like security deployment, interactive design, et cetera. You are a busy guy. >> I am, yes. (laughing) >> One of the things we have been talking about with our guests on the show today, and a great topic that was in the general session was about the value of data, and how do businesses transform to digital businesses. The value in that data has to be critical. I'd love to get your take on as businesses have to leverage that data to become more successful or to become successful as digital businesses, we know the security of the perimeter is not the only thing. It needs to be with the data. What is Azure doing to secure the cloud for your customers, and how do you help them mitigate or deal with the proliferation of mobile devices and IOT devices that they have that are connecting to their networks? >> Digital disruption is affecting everybody, and it is a huge thing that many companies are struggling to understand and to adopt to their business models, and to really leverage what digital can do for them, and certainly we are doing in the public cloud with Azure helps that significantly. As you mentioned, there is just a proliferation of devices, a proliferation of data, so how do you have defense in depth so you don't have perimeter-based security, but you actually have defense in depth at every level, and at its heart, it really falls down to how do you do encryption at rest, how do you secure the data encrypted? Who holds the keys for the data? What is the proliferation of the keys? How did the controls manage for that? Of course, of the data is encrypted, you really want to be able to do things upon it. You want to be able computer over it. You want to be able to queries, analytics, everything, so there's the question of how to securely exchange the keys? How do you make sure that the right virtual machines are running, the right computers running at the time to do the queries? That's the set of controls and security models and services that we provide in Azure that makes it super easy for customers to actually use that. >> Azure represent what's called the second big transformation for Microsoft where the first one might have been associated with Explorer, those amazing things that Microsoft did to transform itself in the 1990s and it seems to be going pretty well. How is security facilitating this transformation from a customer value proposition? >> Security is absolutely the number one question that every customer has whenever they start talking about the cloud, and so we take that very, very seriously. Microsoft spends over billion dollars a year on all of our security products all up. We have literally armies of people who do nothing every day but wake up and make sure that the product is secure, and that really boils down to two big pieces. One is how do we keep the platform secure from the security control that we have ourselves in the compliance ADA stations and everything to make sure that when customers bring their workloads to us, they are in fact kept secure. Second is a set of security controls that we provide the customers so they can actually secure their workloads, integrate their security models with whatever they're running on premise, and have the right security models, ADA stations, multifactor authentication, identity controls, et cetera for their own workloads. >> Security is very context specific. I'm not necessarily getting into a conversation about industry or whatnot, but in terms of the classifications of services that need to be provided, we were talking a little bit about how some of the services that you provide end being part of the architecture for other services within the Azure cloud. Talk a little bit about how you envision security over time evolving as a way of thinking about how different elements of the cloud are going to be integrated and come together in the role that security is going to play in making that possible and easy. >> You are absolutely right. Azure is composed of, right now, 80 some-odd different services and there's definitely a layering where for example, my components around the compute pieces are used by the higher order of services around HD insight and some of the analytic services and such, and so the security models we have in place internally for compute in turn are used by those higher order services, and the real value we can provide is having a common customer-facing security model for customers, so there is a common way by which they can access the control plane, do management operations upon these services, how they can access the endpoints of the services using a common identity model, a common security model, role-based access control, again, from a common perspective, logging, auditing, reporting, so all this has to be cohesive, correct, and unified so that customers aren't facing this tumultuous array of different services that speak different languages, so to speak. >> We are here at Fortinet Accelerate 2017. Tell us how long Microsoft Azure and Fortinet have been working together, and what are you most excited about with some of the announcements from Fortinet today? >> Microsoft and Fortinet partnership has been going on for quite some time. Specifically in Azure space we've been doing two different, two major thrusts around integration with the Azure Security Center which is a set of services that we have within Azure that provides turnkey access to many, many different vendors including Fortinet as one of our primary partners, and Fortinet also has all their products in Azure marketplace so that customers can readily in a turnkey manner use Fortinet next generation firewalls and such as virtual machines, incorporate those directly within their workloads, and have a very seamless billing model, a very seamless partnership model, a very seamless go-to-market strategy for how we jointly promote, jointly provide the services. >> One of the things that one of our guests was talking with us about today was really about it's an easy sell, if you will, at the C-level to sell the value of investing in the right infrastructure to secure environments. Looking at that in correlation to the fact that there's always historically been a challenge or concerned with security when it comes to enterprises moving workloads to the cloud, I'm curious about this easy-sell position that cyber security and the rise of attacks brings to seeing the adoption of more enterprise workloads. We are seeing numbers that are going to show, or predicting that north of 85% of enterprise workloads will be in the cloud by 2020. How much is Microsoft Azure seeing the fact that cyber security attacks are becoming more and more common, hitting some pretty big targets, affecting a lot of big names. How much are using that as an impetus to and maybe drive that adoption higher and higher from an enterprise perspective? >> Absolutely, I see that everyday. I give many, many talks to the C-level, to CSOs, CEOs, et cetera, and I can say in many industries like the banking industry, financial sector, 18 months ago banks did not have any interest in public cloud. Is just like, "Thank you, we have no interest in cloud," but recently there has been the dawning realization that Azure and the public cloud products are in fact, in many cases, more secure than what the banks and other financial industry sectors can actually provide themselves because we are providing huge amounts of investments from an ongoing basis that we can actually provide better security, better integrated security than what they can afford on premise, so as a result, we are seeing this now, literally, stampede of customers coming to us and saying, "Okay, I get it. "You can actually have a very, very "highly secure environment. "You can provide security controls "that can go well above and beyond "whatever I could do on premise, "and it's better integrated "than what I could ever pull together on premise." >> One of the reasons for that is because of the challenge of finding talent, and you guys can find a really talented person, bring them in, and that person can build security architectures for your cloud that then can be served, can be used by a lot of different customers, so what will be the role of or how will this need for talent in the future, what would be the role for how people engage your people, client's people engage your people to ensure that that people side and moves forward, and how do you keep scaling that is you scale the cloud? >> Certainly people are always the bottleneck in virtually every industry, and specifically within the computing space. The value that we are seeing from customers is that the people that they had previously on premise who were working to secure the base level common infrastructure are now freed because they don't have to do that work. They can do other interesting things at the application level and move their value added further up the stack which means I can innovate more rapidly, they can add more features more quickly, because they are not having to worry about the lower-level infrastructure pieces that are secured by Azure, so we are seeing the dawning realization that we are moving to this new golden age where there is higher degree of agility with respect to innovation happening at the application level, because remember, applications have to be, if you are having a compliant workload, if you are having PCI compliance within the credit card industry for example, you have to have the entire application and its infrastructure part of the compliance boundary, so that means when you are building that app, you have to give your auditors the complete stack for them to pass that. If you are only having to worry about this much as opposed to that much, then the amount of work that you can do, the amount of integration, the amount agility, the amount of innovation you can do at that level is many orders of magnitude higher, so you really see that the value that a lot of customers are having here is that their talented people can be put to use on more important higher order business-related problems as opposed to lower-level infrastructure level issues. >> Let's talk about that for second because one of the things that we see within our research is that the era of cloud as renting virtual machines is starting to transition as people start renting applications, or applications as services that they themselves can start putting together. Partly the reason why that's exciting is because it will liberate more developers. It brings more developers into the process of creating value in the cloud, but as they do that, they now have visibility, or they are going to be doing things that touch an enormous set of resources, so how do you make security easier to developers in Azure? >> The key is that we can do high degrees of integration at the low level between these very services. >> Peter: It goes back to that issue of a cascading of your stuff up into the other Azure services. >> Absolutely, I mean think about it, we sat on top a mountain of information. We have analytics and log files that know about virtually everything that's happening in the cloud, and we can have machine learning, we can have intelligence, we can have machine intelligence and such, that can extract signals from noise that would otherwise be impossible to discover from a single customer's perspective. If you have a low and slow attack by some sort of persistent individual, the fact that they are trying the slow and low attack means that we are able to pull that signal out and extract that information that would not be really physically possible, or economically possible for most companies to do on premise. >> Does this get embedded to some of the toolkits that we are going to use to build these next-generation cloud-based apps? >> It gets embedded into the toolkits, but it also gets embedded at the set of services like the Azure Security Center. A single pane of glass that's integrated with the products from Fortinet and others where the customer can go and have a single view across all their work was running within Azure and get comprehensive alerts and understanding about the analytics that we are able to pull out and provide to those customers. >> What's next? >> Security is an ever evolving field, and the bad guys are always trying new things, so the work that is really happening, a lot of the innovation that's happening is within the analytics, machine learning space around being able to pull more log files out, being able to refine the algorithms and basically being able to provide more AI to the logs themselves so that we can provide integrated alerts, like for example, if you have a kill chain of an individual coming in attacking one of your product, and then using that to the lateral mobility to other products, or other services within your product, we can pull this together in a common log. We can show to customers here's the sequence of this one individual that across three, or four, or five different services. You have top level disability, and we can give you then guidance to say if you insert separation of duties between these two individuals, then you could've broken that kill chain. We can do proactive guidance to customers to help them secure their own workloads even if they necessarily initially were not deployed in a necessarily most secure manner. >> George, we just have a couple of minutes left, but I'd like to get your perspective. You showed a tremendous amount of the accomplishments that Azure has made in public cloud and in security. What are the opportunities for partners to sell and resell Azure services? >> Absolutely. Microsoft has historically always worked incredibly well with partners. We have a very large partner ecosystem. >> Peter: It's the biggest. >> Is the biggest, exactly. Okay, I don't want to brag too much, yes. (laughing) >> That's what I'm here for, George. >> We see specifically in the security space that partners are increasingly, around 40% of their revenue increasingly is coming from cloud-based assets, cloud-based sales. We are setting up the necessary partner channels and partner models where we can make sure that the reseller channels and our partners are an integral part of our environment, and they can get the necessary revenue shares, and we can give them the leads on how the whole system evolves. Absolutely we believe that partners are first and foremost to our success, and we are making deep, deep, deep investments in the partner programs to make that possible. >> Well George, we wish you and Microsoft Azure continued success as well as your partnership with Fortinet. We thank you so much for taking the time to join us on theCUBE today. >> Thank you. >> And for my cohost, Peter Burris, I'm Lisa Martin. Stick around, we will be right back on theCUBE.

but cute at servicenow knowledge 14 is sponsored by service now here are your hosts Dave vellante and Jeff Creek we're back hi everybody this is Dave vellante with Jeff Frick we're here live at knowledge 14 this is service now it's big customer event about 6,600 people up from about four thousand last year as we've been saying it's kind of tracking the growth of service now which has been pretty meteoric we heard from Mike scarpelli the CFO Frank's loot men they're really doubling down and it's exciting to see we're here in San Francisco where all the action is Jeffrey Moore is here author consultant pundits all-around smart guy cube alum greatly again thank you here so um so you're speaking at the CIO decisions i love the fact that they got so many CIOs here who real CIA a lot of times these conferences you get to you know the infrastructure guys but so what's the vibe like over there well you know it's kind of cool because if you think about service now and you go back to say 10 years this was all about how to make IT more productive around the ITIL model and you know and you'd use these automated services to do this stuff what's happening and Frank nailed it in the keynote he said look this infrastructure can be turned inside out and you can service enable the entire enterprise not just IT need a service enterprise you know HR you can decision a marketing eight-day any other shared service you can turn into a bunch of services that you can sort of call in and use service now as a platform so so the cios it was all about well that that's a different that's a different vision and so how do we map from the old way of sort of thinking about this is an internal productivity facility to this new way of saying no this is an enterprise enablement platform that's a big that's a big move a little bit like Salesforce going to force calm that same flavor yes sir frank's keynote was talking about how the CIO has to become you know more business savvy and of course we've heard that a lot for years and years and years but in fact a number of the folks that we've had on here at the service now are actually of that hill maybe they came from the business but most CIOs didn't necessarily come from the business they weren't P&L managers they weren't running sales do you see that changing yeah I think what happened in the 20th century was IT was sufficiently complex that frankly you had to be a technical person to do it it just it was just really hard and and yes you needed business consultants but the end of the day you needed ten percent business consultants and ninety percent technical people I think we've come a long way since then in the next generation of stuff is more around systems of engagement these things that that communicate with each other as opposed to systems of record and so the profile the winning IT strategy is migrating from help us run information about our business in the back office to help us actually re-engineer the dynamics of our business in the world in the present and that's like going from from data to behavior them it's a big we call it going from systems of record to systems of engagement it's a big show and is that that transition in your mind is very disruptive so what happens to all those purveyors and buyers of systems of engagement to they morph into obsessive record do they morph into systems of engagement do they just get blown away no it's interesting so so so first of all you're never going to get rid of your systems of record but at the margin we've probably extracted most of the lifetime value from that investment already so you need to maintain them and so the industry is consolidating a round of an anchor set of vendors who we trust to do that but the growth is going to be like if you look at systems of engagement we might have gotten five percent of the lifetime value there so at their margin if you have a dollar to spend people want to spend it in there so the challenge of being an incumbent is I'm not going to lose my base but man the growth is happening over here so the real challenge for that for the incumbent vendors is how can i participate in the new world and still maintain my relationships in the old world whereas the new guys are just coming and saying i don't i'll leave the old world of you guys i just want to play over here i can get your take on the structure of the IT business is you've observed as have i sort of these disruptions and these changes over time so obviously we went from being framed at pc you saw that the competitive line started to get more disintegrated yes i could use that that term a competition occurred on those I see that Intel's ascendancy in Microsoft and Oracle the best database companies the emc was the storage company and everything was sort of you know siloed and but leadership the leadership matrix has largely stayed intact I mean even IBM and okay HP said its ends up and down but it's largely stayed intact do you see the cloud changing that fundamentally changing the economic yes I think yes I think what happened is so in the client server error we did we built the stack what you're just described and every layer of the stack had a leader now I think since 2000 y2k that stack is being compressed meaning there are fewer and fewer vendors that are still in the in that in that leadership cadre and as we go to like cloud and computing the service you start saying well yeah i still have cisco in there i still have IBM in there but maybe i'm buying them as a service rather than as a set of equipment so you kind of can feel that world just I think compressing this look is the right word and where is the experimentation the opportunity to sort of find new places to go to it's very much in this world outboard of the IT data center where it it is about engaging engaging with your customer engaging with your employee engaging with your supply chain and using mobile things and social and you know analytics and cloud and all these new technologies the freedom to do that is is actually outboard of the of the old style I show you what you described as sort of an oligopoly and you've got these big whales and I've always asking you know guys who follow this it are we going to see somebody to disrupt that Amazon is the obvious you have to go to them a three billion dollar you know company growing at sixty percent a year with marginal economics of services that look like software yep but at the same time it's okay they've got this huge lead but it doesn't just make sense to me that it's sustainable I mean because hardware economics never will go to 0 so you would think that somebody was almost like the IBM early pc days remember IBM heavily yep we're domin to play that's kind of what kind of way amazon is now do you do you see that you see more competition from amazon why is it that they don't have direct competition so the less of the last book i wrote in the last the thing i've been working on most recently is around why is it so hard for the established incumbents to catch the next wave and the problem is so you look at why amazon's why is Amazon so unopposed in many of its initiatives well their business model in the economic model is completely divorced from the incumbent model and so you look at the incumbent in there going it's not that I don't see what the guys are doing I get what they're doing I just don't see how I can get my investors or my my whole infrastructure on to that new place in my example that was code at so you know Antonio Perez came from HP he knew what he was getting into he understood digital everybody at Kodak understood digital but they couldn't get to the other place so in this it would call it escape velocity how do you free yourself from your own paths and you you really do have to take a pretty dramatic approach to it and I think by the way i think i'm looking at microsoft in particular i think it I think Microsoft's going to give a very very big run at doing it and but I think that they're still more the exception than the rule you would wish that every one of those vendors would say look you know because every CIO here if any of those vendors came to him and said hey we're going to really try to play here will you help they'd say yes they don't want to change their relationships but but we get trapped in these business models and then you sort of grind and you grind and grind and after a while it's like well man you've just ground yourself to do I owe the classic label Christensen right individuals dilemma and it also makes a question is d said David's been the same characters kind of changing companies had not Jeff Bezos and Amazon come in with a completely different model to drive cloud with the other people who still has to transfer so they want to give credit to you want to bet it to be so so you want to give credit to Benioff by the way Benioff has been has been the kind of prow of a ship that brings in the illusory at work day brings in netsuite brings in service service now you know so the software-as-a-service thing is coming in at one level and remember if you were an on-premise guy it's very very how many years did did SI p commit an enormous amount of money to say we're going to have a great cloud offering and it just it's so hard so so it is so and then you're looking now at this sort of this next layer of collaborative IT and you're seeing box and octant hang all these cool thing and analytics and splunk consumer logic and all these companies going really I mean I you know I mean if your fear of my age is like okay you have a t-shirt they got love to you think I'm a teacher but but but the point is this free space and they're saying there's these cool problems to solve we're not encumbered by any of the legacy we're going to race ahead and so if you're a CIO well we spent most of our time with the cios today was ok i have established set of relationships here i'm not going to abandon them but at the margin i need them to help me think about the future I thought these really start sparkly new startups some i'm sure not going to exist next year but some are going to be the leaders so how play that game right now and and the pressure it's putting on the IT organization is the people I know that are good at this are not the people that are good at this and so how do I so we had to talk about talent and how do you manage and how do you create career paths and and is it or do you have a infrastructure officer vs an innovation office I'm it was all around that same prob right and then oh by the way there's Hadoop and mobile and big data and some of these other just open source innovations that are being just thrown all these guys played it is so from a technology plate from a technology play if you're technologists it's like bring it on right but I think the interesting thing is and most of my career aighty was about the business so you ran a business and you had IT systems which gave you information about your business what's happened in the last 15 years is that more and more sectors of the economy i T is becoming the business so you saw what happened the newspapers in facilitate with IT isn't about the newspaper business IT is displacing the newspaper business Google is displaced in the media business amazon is displacing retail you know mobile banking is displacing banking Airbnb uber I mean this so there we have the taxi guys are worried them it and so you start saying it isn't IT isn't about the business it's a digital world and and so all of us and that was it i think that was probably at the core of the discussion so which cio am i what do I have permission to be would do my colleagues get this you know am I competent to do it if they do I mean you've talked about this a lot and you've given a number of examples so so was nicked car just dead wrong in 2003 or just to a narrow it is to keep what he was saying I believe is that systems of record okay are dead I think at that time by the way it wasn't obvious there was anything else because it no serious i can remember to you know the whole venture community kind of abandoned itv4 about researcher ivan on 101 yeah it was and even in the end even in the physical infrastructure there's still the idea is the basis of the competitive and about the reporting system yeah and i think this issue about so i think there's still a few businesses we're really IT still is about the business and you know what you can kind of stick with whatever you were doing you'll be okay but if your business is under an existential threat meaning the new IT model eviscerates your business model which arguably you could say all those both those incumbent stack vendors you know I mean cloud does eviscerate the on-premise hardware data center business model which was the fundamental foundation of IT as I knew it for all my business career and now all this it's like holy how do i how do i how do I deal with it so we talk about Amazon as a potential you know new you know big whale Salesforce is obviously he's got it but they've been around since 99 there's going to be exception mm-hmm proves the rule I don't maybe a service now or a workday you know we'll see if this market is big enough it looks like it it might be what often happens is they these guys let's get gobbled up or Larry Ellison writes a check you say these to denigrate people who write write checks not code I think the biggest matter and they got such mass never was afraid to reinvent himself change the game change the dynamics of the industry so do you think we will see a another big player and where will that comfort will it be the SAS guys will it be the sum of the guys out of the hadoop world what I don't think it will so here here's what I don't think will work I don't think you can be an established incumbent vendor under this compression power and write a check and get yourself back I think what happens when you write a check if you just bring a hot property into cold molecules and it loses its exactly exactly so I don't think that will work I think if you want to be one of these incumbents and succeed over here you have to actually pull part of your own DNA and capability and we literally just jump and then I think you can acquire it to it to build a thing there but what Larry did was he consolidate he basically was the first guy to figure out Nick Carr is right I need to buy up all the properties yep and brother George ball and run a maintenance business which by the way came to read and Georgia computer associates had that play up in the eighties it's the same play with this is a different plan well I love what you say in emc is an interesting one to watch the way to chi is setting up this Federation with pivotal and VMware you know who see we'll see what happens with the quarry NC and I think VI 3 of 8 yeah I think that that is I mean VMware's one of the wonderful examples of think we're a company did not cause the hot molecules become the cold molecules the thing you wonder there though is it feels a little bit like a like a holding company if you will and so and by the way vmware is in a curious tweener right like they kind of were the most they made the old stack incredibly productive so in some sense they can feel like they're part of the old world right they're probably the newest kid on the old world but then you think well yeah but I want to look at their plan now they want to be into software-defined networks they wanted me to software-defined data centers they definitely want to play over here and what it's in this case so state partners Wow one could argue that that was it because of what big in the cloud virtualize computing absolutely absolutely so what're you working on these days that's exciting well so that I think this issue of working with management teams to say okay look this is a self-imposed exile that we're putting ourselves under you know we get it i'll call it the Kodak problem because I don't want to talk about anybody in high tech specifically at the moment but the point is every management team in the established vendor group puts itself on a self imposed discipline to make you know certain kinds of eps things certain kinds of growth you know whatever it is the expectations of their investors and you look at the situation you say guys that is a slope glide path to extinction we all know that and by the way off the record they know it's no it's not that that is this is not a failure of it like this is a failure of will so then the question is well so how do you negotiate a different path and part of it is you have to make you have you have to be able to tell a story of your investors part of it is you have to negotiate a different operating model inside the company and what they've done so far is they said well okay we've got our established businesses and we've got our innovative businesses and we know enough to keep them apart so that part is not the problem and they actually come up with cool stuff the the moment of truth is when can you scale any of these innovative businesses to compete to actually be a material part of your historical portfolio meaning in my terminology at least ten percent of your total revenue going to twenty percent in what happens in that journey is it a key point you have to draw on the resources of your established business and all the people that make their living and they're compensated on getting the next quarter in the next quarter go guys I can't make the quarter and do this and you've got it you've got to find a way to say you know if we don't figure out a way to pull some of that resource over here and play our next hand will invent everything in the world but we'll never get it to scale and so there's there's a bunch of stuff around business model planning and then Investor Relations organizational development it's all around saying and the key there's two key ideas idea number one is it's a go-to-market problem not an RD problem you do not have an innovation problem you can't get your thing to market and the second cool idea is you can only do one of the time and everybody says well but give have the risk to so high you got a three or four or five of these things maybe want to work it's like know the sacrifice is so great if you put two or more horses in the race people people won't even run so the other one that's a focus and don't it's ok not to make the quarter that's like on American looking like michael dunn right i mean that's obsessively what he's hoping to be able to do and i think one of the reasons you see people go private is to say i can't play this game bye-bye normal public company protocol i mean i like to but i can't get there from here now i actually don't think every company ought to have to go private to do this but i think they do have to change their playboys all right Jeff we have to leave it there hey great to see you thank you very much me feel smarter just hanging out with you right there buddy we'll be right back after this is the cube you

(chill electronic music) >> Hello everyone, welcome to theCUBE's presentation of the AWS Startups Showcase, AI and machine learning, the top startups building generative AI on AWS. This is the season three, episode one of the ongoing series covering the exciting startups from the AWS ecosystem, talk about AI and machine learning. Can't believe it's three years and season one. I'm your host, John Furrier. Got a great guest today, we're joined by Joseph Nelson, the co-founder and CEO of Roboflow, doing some cutting edge stuff around computer vision and really at the front end of this massive wave coming around, large language models, computer vision. The next gen AI is here, and it's just getting started. We haven't even scratched a service. Thanks for joining us today. >> Thanks for having me. >> So you got to love the large language model, foundation models, really educating the mainstream world. ChatGPT has got everyone in the frenzy. This is educating the world around this next gen AI capabilities, enterprise, image and video data, all a big part of it. I mean the edge of the network, Mobile World Conference is happening right now, this month, and it's just ending up, it's just continue to explode. Video is huge. So take us through the company, do a quick explanation of what you guys are doing, when you were founded. Talk about what the company's mission is, and what's your North Star, why do you exist? >> Yeah, Roboflow exists to really kind of make the world programmable. I like to say make the world be read and write access. And our North Star is enabling developers, predominantly, to build that future. If you look around, anything that you see will have software related to it, and can kind of be turned into software. The limiting reactant though, is how to enable computers and machines to understand things as well as people can. And in a lot of ways, computer vision is that missing element that enables anything that you see to become software. So in the virtue of, if software is eating the world, computer vision kind of makes the aperture infinitely wide. It's something that I kind of like, the way I like to frame it. And the capabilities are there, the open source models are there, the amount of data is there, the computer capabilities are only improving annually, but there's a pretty big dearth of tooling, and an early but promising sign of the explosion of use cases, models, and data sets that companies, developers, hobbyists alike will need to bring these capabilities to bear. So Roboflow is in the game of building the community around that capability, building the use cases that allow developers and enterprises to use computer vision, and providing the tooling for companies and developers to be able to add computer vision, create better data sets, and deploy to production, quickly, easily, safely, invaluably. >> You know, Joseph, the word in production is actually real now. You're seeing a lot more people doing in production activities. That's a real hot one and usually it's slower, but it's gone faster, and I think that's going to be more the same. And I think the parallel between what we're seeing on the large language models coming into computer vision, and as you mentioned, video's data, right? I mean we're doing video right now, we're transcribing it into a transcript, linking up to your linguistics, times and the timestamp, I mean everything's data and that really kind of feeds. So this connection between what we're seeing, the large language and computer vision are coming together kind of cousins, brothers. I mean, how would you compare, how would you explain to someone, because everyone's like on this wave of watching people bang out their homework assignments, and you know, write some hacks on code with some of the open AI technologies, there is a corollary directly related to to the vision side. Can you explain? >> Yeah, the rise of large language models are showing what's possible, especially with text, and I think increasingly will get multimodal as the images and video become ingested. Though there's kind of this still core missing element of basically like understanding. So the rise of large language models kind of create this new area of generative AI, and generative AI in the context of computer vision is a lot of, you know, creating video and image assets and content. There's also this whole surface area to understanding what's already created. Basically digitizing physical, real world things. I mean the Metaverse can't be built if we don't know how to mirror or create or identify the objects that we want to interact with in our everyday lives. And where computer vision comes to play in, especially what we've seen at Roboflow is, you know, a little over a hundred thousand developers now have built with our tools. That's to the tune of a hundred million labeled open source images, over 10,000 pre-trained models. And they've kind of showcased to us all of the ways that computer vision is impacting and bringing the world to life. And these are things that, you know, even before large language models and generative AI, you had pretty impressive capabilities, and when you add the two together, it actually unlocks these kind of new capabilities. So for example, you know, one of our users actually powers the broadcast feeds at Wimbledon. So here we're talking about video, we're streaming, we're doing things live, we've got folks that are cropping and making sure we look good, and audio/visual all plugged in correctly. When you broadcast Wimbledon, you'll notice that the camera controllers need to do things like track the ball, which is moving at extremely high speeds and zoom crop, pan tilt, as well as determine if the ball bounced in or out. The very controversial but critical key to a lot of tennis matches. And a lot of that has been historically done with the trained, but fallible human eye and computer vision is, you know, well suited for this task to say, how do we track, pan, tilt, zoom, and see, track the tennis ball in real time, run at 30 plus frames per second, and do it all on the edge. And those are capabilities that, you know, were kind of like science fiction, maybe even a decade ago, and certainly five years ago. Now the interesting thing, is that with the advent of of generative AI, you can start to do things like create your own training data sets, or kind of create logic around once you have this visual input. And teams at Tesla have actually been speaking about, of course the autopilot team's focused on doing vision tasks, but they've combined large language models to add reasoning and logic. So given that you see, let's say the tennis ball, what do you want to do? And being able to combine the capabilities of what LLM's represent, which is really a lot of basically, core human reasoning and logic, with computer vision for the inputs of what's possible, creates these new capabilities, let alone multimodality, which I'm sure we'll talk more about. >> Yeah, and it's really, I mean it's almost intoxicating. It's amazing that this is so capable because the cloud scales here, you got the edge developing, you can decouple compute power, and let Moore's law and all the new silicone and the processors and the GPUs do their thing, and you got open source booming. You're kind of getting at this next segment I wanted to get into, which is the, how people should be thinking about these advances of the computer vision. So this is now a next wave, it's here. I mean I'd love to have that for baseball because I'm always like, "Oh, it should have been a strike." I'm sure that's going to be coming soon, but what is the computer vision capable of doing today? I guess that's my first question. You hit some of it, unpack that a little bit. What does general AI mean in computer vision? What's the new thing? Because there are old technology's been around, proprietary, bolted onto hardware, but hardware advances at a different pace, but now you got new capabilities, generative AI for vision, what does that mean? >> Yeah, so computer vision, you know, at its core is basically enabling machines, computers, to understand, process, and act on visual data as effective or more effective than people can. Traditionally this has been, you know, task types like classification, which you know, identifying if a given image belongs in a certain category of goods on maybe a retail site, is the shoes or is it clothing? Or object detection, which is, you know, creating bounding boxes, which allows you to do things like count how many things are present, or maybe measure the speed of something, or trigger an alert when something becomes visible in frame that wasn't previously visible in frame, or instant segmentation where you're creating pixel wise segmentations for both instance and semantic segmentation, where you often see these kind of beautiful visuals of the polygon surrounding objects that you see. Then you have key point detection, which is where you see, you know, athletes, and each of their joints are kind of outlined is another more traditional type problem in signal processing and computer vision. With generative AI, you kind of get a whole new class of problem types that are opened up. So in a lot of ways I think about generative AI in computer vision as some of the, you know, problems that you aimed to tackle, might still be better suited for one of the previous task types we were discussing. Some of those problem types may be better suited for using a generative technique, and some are problem types that just previously wouldn't have been possible absent generative AI. And so if you make that kind of Venn diagram in your head, you can think about, okay, you know, visual question answering is a task type where if I give you an image and I say, you know, "How many people are in this image?" We could either build an object detection model that might count all those people, or maybe a visual question answering system would sufficiently answer this type of problem. Let alone generative AI being able to create new training data for old systems. And that's something that we've seen be an increasingly prominent use case for our users, as much as things that we advise our customers and the community writ large to take advantage of. So ultimately those are kind of the traditional task types. I can give you some insight, maybe, into how I think about what's possible today, or five years or ten years as you sort go back. >> Yes, definitely. Let's get into that vision. >> So I kind of think about the types of use cases in terms of what's possible. If you just imagine a very simple bell curve, your normal distribution, for the longest time, the types of things that are in the center of that bell curve are identifying objects that are very common or common objects in context. Microsoft published the COCO Dataset in 2014 of common objects and contexts, of hundreds of thousands of images of chairs, forks, food, person, these sorts of things. And you know, the challenge of the day had always been, how do you identify just those 80 objects? So if we think about the bell curve, that'd be maybe the like dead center of the curve, where there's a lot of those objects present, and it's a very common thing that needs to be identified. But it's a very, very, very small sliver of the distribution. Now if you go out to the way long tail, let's go like deep into the tail of this imagined visual normal distribution, you're going to have a problem like one of our customers, Rivian, in tandem with AWS, is tackling, to do visual quality assurance and manufacturing in production processes. Now only Rivian knows what a Rivian is supposed to look like. Only they know the imagery of what their goods that are going to be produced are. And then between those long tails of proprietary data of highly specific things that need to be understood, in the center of the curve, you have a whole kind of messy middle, type of problems I like to say. The way I think about computer vision advancing, is it's basically you have larger and larger and more capable models that eat from the center out, right? So if you have a model that, you know, understands the 80 classes in COCO, well, pretty soon you have advances like Clip, which was trained on 400 million image text pairs, and has a greater understanding of a wider array of objects than just 80 classes in context. And over time you'll get more and more of these larger models that kind of eat outwards from that center of the distribution. And so the question becomes for companies, when can you rely on maybe a model that just already exists? How do you use your data to get what may be capable off the shelf, so to speak, into something that is usable for you? Or, if you're in those long tails and you have proprietary data, how do you take advantage of the greatest asset you have, which is observed visual information that you want to put to work for your customers, and you're kind of living in the long tails, and you need to adapt state of the art for your capabilities. So my mental model for like how computer vision advances is you have that bell curve, and you have increasingly powerful models that eat outward. And multimodality has a role to play in that, larger models have a role to play in that, more compute, more data generally has a role to play in that. But it will be a messy and I think long condition. >> Well, the thing I want to get, first of all, it's great, great mental model, I appreciate that, 'cause I think that makes a lot of sense. The question is, it seems now more than ever, with the scale and compute that's available, that not only can you eat out to the middle in your example, but there's other models you can integrate with. In the past there was siloed, static, almost bespoke. Now you're looking at larger models eating into the bell curve, as you said, but also integrating in with other stuff. So this seems to be part of that interaction. How does, first of all, is that really happening? Is that true? And then two, what does that mean for companies who want to take advantage of this? Because the old model was operational, you know? I have my cameras, they're watching stuff, whatever, and like now you're in this more of a, distributed computing, computer science mindset, not, you know, put the camera on the wall kind of- I'm oversimplifying, but you know what I'm saying. What's your take on that? >> Well, to the first point of, how are these advances happening? What I was kind of describing was, you know, almost uni-dimensional in that you have like, you're only thinking about vision, but the rise of generative techniques and multi-modality, like Clip is a multi-modal model, it has 400 million image text pairs. That will advance the generalizability at a faster rate than just treating everything as only vision. And that's kind of where LLMs and vision will intersect in a really nice and powerful way. Now in terms of like companies, how should they be thinking about taking advantage of these trends? The biggest thing that, and I think it's different, obviously, on the size of business, if you're an enterprise versus a startup. The biggest thing that I think if you're an enterprise, and you have an established scaled business model that is working for your customers, the question becomes, how do you take advantage of that established data moat, potentially, resource moats, and certainly, of course, establish a way of providing value to an end user. So for example, one of our customers, Walmart, has the advantage of one of the largest inventory and stock of any company in the world. And they also of course have substantial visual data, both from like their online catalogs, or understanding what's in stock or out of stock, or understanding, you know, the quality of things that they're going from the start of their supply chain to making it inside stores, for delivery of fulfillments. All these are are visual challenges. Now they already have a substantial trove of useful imagery to understand and teach and train large models to understand each of the individual SKUs and products that are in their stores. And so if I'm a Walmart, what I'm thinking is, how do I make sure that my petabytes of visual information is utilized in a way where I capture the proprietary benefit of the models that I can train to do tasks like, what item was this? Or maybe I'm going to create AmazonGo-like technology, or maybe I'm going to build like delivery robots, or I want to automatically know what's in and out of stock from visual input fees that I have across my in-store traffic. And that becomes the question and flavor of the day for enterprises. I've got this large amount of data, I've got an established way that I can provide more value to my own customers. How do I ensure I take advantage of the data advantage I'm already sitting on? If you're a startup, I think it's a pretty different question, and I'm happy to talk about. >> Yeah, what's startup angle on this? Because you know, they're going to want to take advantage. It's like cloud startups, cloud native startups, they were born in the cloud, they never had an IT department. So if you're a startup, is there a similar role here? And if I'm a computer vision startup, what's that mean? So can you share your your take on that, because there'll be a lot of people starting up from this. >> So the startup on the opposite advantage and disadvantage, right? Like a startup doesn't have an proven way of delivering repeatable value in the same way that a scaled enterprise does. But it does have the nimbleness to identify and take advantage of techniques that you can start from a blank slate. And I think the thing that startups need to be wary of in the generative AI enlarged language model, in multimodal world, is building what I like to call, kind of like sandcastles. A sandcastle is maybe a business model or a capability that's built on top of an assumption that is going to be pretty quickly wiped away by improving underlying model technology. So almost like if you imagine like the ocean, the waves are coming in, and they're going to wipe away your progress. You don't want to be in the position of building sandcastle business where, you don't want to bet on the fact that models aren't going to get good enough to solve the task type that you might be solving. In other words, don't take a screenshot of what's capable today. Assume that what's capable today is only going to continue to become possible. And so for a startup, what you can do, that like enterprises are quite comparatively less good at, is embedding these capabilities deeply within your products and delivering maybe a vertical based experience, where AI kind of exists in the background. >> Yeah. >> And we might not think of companies as, you know, even AI companies, it's just so embedded in the experience they provide, but that's like the vertical application example of taking AI and making it be immediately usable. Or, of course there's tons of picks and shovels businesses to be built like Roboflow, where you're enabling these enterprises to take advantage of something that they have, whether that's their data sets, their computes, or their intellect. >> Okay, so if I hear that right, by the way, I love, that's horizontally scalable, that's the large language models, go up and build them the apps, hence your developer focus. I'm sure that's probably the reason that the tsunami of developer's action. So you're saying picks and shovels tools, don't try to replicate the platform of what could be the platform. Oh, go to a VC, I'm going to build a platform. No, no, no, no, those are going to get wiped away by the large language models. Is there one large language model that will rule the world, or do you see many coming? >> Yeah, so to be clear, I think there will be useful platforms. I just think a lot of people think that they're building, let's say, you know, if we put this in the cloud context, you're building a specific type of EC2 instance. Well, it turns out that Amazon can offer that type of EC2 instance, and immediately distribute it to all of their customers. So you don't want to be in the position of just providing something that actually ends up looking like a feature, which in the context of AI, might be like a small incremental improvement on the model. If that's all you're doing, you're a sandcastle business. Now there's a lot of platform businesses that need to be built that enable businesses to get to value and do things like, how do I monitor my models? How do I create better models with my given data sets? How do I ensure that my models are doing what I want them to do? How do I find the right models to use? There's all these sorts of platform wide problems that certainly exist for businesses. I just think a lot of startups that I'm seeing right now are making the mistake of assuming the advances we're seeing are not going to accelerate or even get better. >> So if I'm a customer, if I'm a company, say I'm a startup or an enterprise, either one, same question. And I want to stand up, and I have developers working on stuff, I want to start standing up an environment to start doing stuff. Is that a service provider? Is that a managed service? Is that you guys? So how do you guys fit into your customers leaning in? Is it just for developers? Are you targeting with a specific like managed service? What's the product consumption? How do you talk to customers when they come to you? >> The thing that we do is enable, we give developers superpowers to build automated inventory tracking, self-checkout systems, identify if this image is malignant cancer or benign cancer, ensure that these products that I've produced are correct. Make sure that that the defect that might exist on this electric vehicle makes its way back for review. All these sorts of problems are immediately able to be solved and tackled. In terms of the managed services element, we have solutions as integrators that will often build on top of our tools, or we'll have companies that look to us for guidance, but ultimately the company is in control of developing and building and creating these capabilities in house. I really think the distinction is maybe less around managed service and tool, and more around ownership in the era of AI. So for example, if I'm using a managed service, in that managed service, part of their benefit is that they are learning across their customer sets, then it's a very different relationship than using a managed service where I'm developing some amount of proprietary advantages for my data sets. And I think that's a really important thing that companies are becoming attuned to, just the value of the data that they have. And so that's what we do. We tell companies that you have this proprietary, immense treasure trove of data, use that to your advantage, and think about us more like a set of tools that enable you to get value from that capability. You know, the HashiCorp's and GitLab's of the world have proven like what these businesses look like at scale. >> And you're targeting developers. When you go into a company, do you target developers with freemium, is there a paid service? Talk about the business model real quick. >> Sure, yeah. The tools are free to use and get started. When someone signs up for Roboflow, they may elect to make their work open source, in which case we're able to provide even more generous usage limits to basically move the computer vision community forward. If you elect to make your data private, you can use our hosted data set managing, data set training, model deployment, annotation tooling up to some limits. And then usually when someone validates that what they're doing gets them value, they purchase a subscription license to be able to scale up those capabilities. So like most developer centric products, it's free to get started, free to prove, free to poke around, develop what you think is possible. And then once you're getting to value, then we're able to capture the commercial upside in the value that's being provided. >> Love the business model. It's right in line with where the market is. There's kind of no standards bodies these days. The developers are the ones who are deciding kind of what the standards are by their adoption. I think making that easy for developers to get value as the model open sources continuing to grow, you can see more of that. Great perspective Joseph, thanks for sharing that. Put a plug in for the company. What are you guys doing right now? Where are you in your growth? What are you looking for? How should people engage? Give the quick commercial for the company. >> So as I mentioned, Roboflow is I think one of the largest, if not the largest collections of computer vision models and data sets that are open source, available on the web today, and have a private set of tools that over half the Fortune 100 now rely on those tools. So we're at the stage now where we know people want what we're working on, and we're continuing to drive that type of adoption. So companies that are looking to make better models, improve their data sets, train and deploy, often will get a lot of value from our tools, and certainly reach out to talk. I'm sure there's a lot of talented engineers that are tuning in too, we're aggressively hiring. So if you are interested in being a part of making the world programmable, and being at the ground floor of the company that's creating these capabilities to be writ large, we'd love to hear from you. >> Amazing, Joseph, thanks so much for coming on and being part of the AWS Startup Showcase. Man, if I was in my twenties, I'd be knocking on your door, because it's the hottest trend right now, it's super exciting. Generative AI is just the beginning of massive sea change. Congratulations on all your success, and we'll be following you guys. Thanks for spending the time, really appreciate it. >> Thanks for having me. >> Okay, this is season three, episode one of the ongoing series covering the exciting startups from the AWS ecosystem, talking about the hottest things in tech. I'm John Furrier, your host. Thanks for watching. (chill electronic music)

(soft music) >> Hello, everyone. Welcome to theCUBE's presentation of the AWS Startup Showcase. AI and Machine Learning: Top Startups Building Foundational Model Infrastructure. This is season 3, episode 1 of the ongoing series covering the exciting stuff from the AWS ecosystem, talking about machine learning and AI. I'm your host, John Furrier and today we are excited to be joined by Luis Ceze who's the CEO of OctoML and Anna Connolly, VP of customer success and experience OctoML. Great to have you on again, Luis. Anna, thanks for coming on. Appreciate it. >> Thank you, John. It's great to be here. >> Thanks for having us. >> I love the company. We had a CUBE conversation about this. You guys are really addressing how to run foundational models faster for less. And this is like the key theme. But before we get into it, this is a hot trend, but let's explain what you guys do. Can you set the narrative of what the company's about, why it was founded, what's your North Star and your mission? >> Yeah, so John, our mission is to make AI sustainable and accessible for everyone. And what we offer customers is, you know, a way of taking their models into production in the most efficient way possible by automating the process of getting a model and optimizing it for a variety of hardware and making cost-effective. So better, faster, cheaper model deployment. >> You know, the big trend here is AI. Everyone's seeing the ChatGPT, kind of the shot heard around the world. The BingAI and this fiasco and the ongoing experimentation. People are into it, and I think the business impact is clear. I haven't seen this in all of my career in the technology industry of this kind of inflection point. And every senior leader I talk to is rethinking about how to rebuild their business with AI because now the large language models have come in, these foundational models are here, they can see value in their data. This is a 10 year journey in the big data world. Now it's impacting that, and everyone's rebuilding their company around this idea of being AI first 'cause they see ways to eliminate things and make things more efficient. And so now they telling 'em to go do it. And they're like, what do we do? So what do you guys think? Can you explain what is this wave of AI and why is it happening, why now, and what should people pay attention to? What does it mean to them? >> Yeah, I mean, it's pretty clear by now that AI can do amazing things that captures people's imaginations. And also now can show things that are really impactful in businesses, right? So what people have the opportunity to do today is to either train their own model that adds value to their business or find open models out there that can do very valuable things to them. So the next step really is how do you take that model and put it into production in a cost-effective way so that the business can actually get value out of it, right? >> Anna, what's your take? Because customers are there, you're there to make 'em successful, you got the new secret weapon for their business. >> Yeah, I think we just see a lot of companies struggle to get from a trained model into a model that is deployed in a cost-effective way that actually makes sense for the application they're building. I think that's a huge challenge we see today, kind of across the board across all of our customers. >> Well, I see this, everyone asking the same question. I have data, I want to get value out of it. I got to get these big models, I got to train it. What's it going to cost? So I think there's a reality of, okay, I got to do it. Then no one has any visibility on what it costs. When they get into it, this is going to break the bank. So I have to ask you guys, the cost of training these models is on everyone's mind. OctoML, your company's focus on the cost side of it as well as the efficiency side of running these models in production. Why are the production costs such a concern and where specifically are people looking at it and why did it get here? >> Yeah, so training costs get a lot of attention because normally a large number, but we shouldn't forget that it's a large, typically one time upfront cost that customers pay. But, you know, when the model is put into production, the cost grows directly with model usage and you actually want your model to be used because it's adding value, right? So, you know, the question that a customer faces is, you know, they have a model, they have a trained model and now what? So how much would it cost to run in production, right? And now without the big wave in generative AI, which rightfully is getting a lot of attention because of the amazing things that it can do. It's important for us to keep in mind that generative AI models like ChatGPT are huge, expensive energy hogs. They cost a lot to run, right? And given that model usage growth directly, model cost grows directly with usage, what you want to do is make sure that once you put a model into production, you have the best cost structure possible so that you're not surprised when it's gets popular, right? So let me give you an example. So if you have a model that costs, say 1 to $2 million to train, but then it costs about one to two cents per session to use it, right? So if you have a million active users, even if they use just once a day, it's 10 to $20,000 a day to operate that model in production. And that very, very quickly, you know, get beyond what you paid to train it. >> Anna, these aren't small numbers, and it's cost to train and cost to operate, it kind of reminds me of when the cloud came around and the data center versus cloud options. Like, wait a minute, one, it costs a ton of cash to deploy, and then running it. This is kind of a similar dynamic. What are you seeing? >> Yeah, absolutely. I think we are going to see increasingly the cost and production outpacing the costs and training by a lot. I mean, people talk about training costs now because that's what they're confronting now because people are so focused on getting models performant enough to even use in an application. And now that we have them and they're that capable, we're really going to start to see production costs go up a lot. >> Yeah, Luis, if you don't mind, I know this might be a little bit of a tangent, but, you know, training's super important. I get that. That's what people are doing now, but then there's the deployment side of production. Where do people get caught up and miss the boat or misconfigure? What's the gotcha? Where's the trip wire or so to speak? Where do people mess up on the cost side? What do they do? Is it they don't think about it, they tie it to proprietary hardware? What's the issue? >> Yeah, several things, right? So without getting really technical, which, you know, I might get into, you know, you have to understand relationship between performance, you know, both in terms of latency and throughput and cost, right? So reducing latency is important because you improve responsiveness of the model. But it's really important to keep in mind that it often leads diminishing returns. Below a certain latency, making it faster won't make a measurable difference in experience, but it's going to cost a lot more. So understanding that is important. Now, if you care more about throughputs, which is the time it takes for you to, you know, units per period of time, you care about time to solution, we should think about this throughput per dollar. And understand what you want is the highest throughput per dollar, which may come at the cost of higher latency, which you're not going to care about, right? So, and the reality here, John, is that, you know, humans and especially folks in this space want to have the latest and greatest hardware. And often they commit a lot of money to get access to them and have to commit upfront before they understand the needs that their models have, right? So common mistake here, one is not spending time to understand what you really need, and then two, over-committing and using more hardware than you actually need. And not giving yourself enough freedom to get your workload to move around to the more cost-effective choice, right? So this is just a metaphoric choice. And then another thing that's important here too is making a model run faster on the hardware directly translates to lower cost, right? So, but it takes a lot of engineers, you need to think of ways of producing very efficient versions of your model for the target hardware that you're going to use. >> Anna, what's the customer angle here? Because price performance has been around for a long time, people get that, but now latency and throughput, that's key because we're starting to see this in apps. I mean, there's an end user piece. I even seeing it on the infrastructure side where they're taking a heavy lifting away from operational costs. So you got, you know, application specific to the user and/or top of the stack, and then you got actually being used in operations where they want both. >> Yeah, absolutely. Maybe I can illustrate this with a quick story with the customer that we had recently been working with. So this customer is planning to run kind of a transformer based model for tech generation at super high scale on Nvidia T4 GPU, so kind of a commodity GPU. And the scale was so high that they would've been paying hundreds of thousands of dollars in cloud costs per year just to serve this model alone. You know, one of many models in their application stack. So we worked with this team to optimize our model and then benchmark across several possible targets. So that matching the hardware that Luis was just talking about, including the newer kind of Nvidia A10 GPUs. And what they found during this process was pretty interesting. First, the team was able to shave a quarter of their spend just by using better optimization techniques on the T4, the older hardware. But actually moving to a newer GPU would allow them to serve this model in a sub two milliseconds latency, so super fast, which was able to unlock an entirely new kind of user experience. So they were able to kind of change the value they're delivering in their application just because they were able to move to this new hardware easily. So they ultimately decided to plan their deployment on the more expensive A10 because of this, but because of the hardware specific optimizations that we helped them with, they managed to even, you know, bring costs down from what they had originally planned. And so if you extend this kind of example to everything that's happening with generative AI, I think the story we just talked about was super relevant, but the scale can be even higher, you know, it can be tenfold that. We were recently conducting kind of this internal study using GPT-J as a proxy to illustrate the experience of just a company trying to use one of these large language models with an example scenario of creating a chatbot to help job seekers prepare for interviews. So if you imagine kind of a conservative usage scenario where the model generates just 3000 words per user per day, which is, you know, pretty conservative for how people are interacting with these models. It costs 5 cents a session and if you're a company and your app goes viral, so from, you know, beginning of the year there's nobody, at the end of the year there's a million daily active active users in that year alone, going from zero to a million. You'll be spending about $6 million a year, which is pretty unmanageable. That's crazy, right? >> Yeah. >> For a company or a product that's just launching. So I think, you know, for us we see the real way to make these kind of advancements accessible and sustainable, as we said is to bring down cost to serve using these techniques. >> That's a great story and I think that illustrates this idea that deployment cost can vary from situation to situation, from model to model and that the efficiency is so strong with this new wave, it eliminates heavy lifting, creates more efficiency, automates intellect. I mean, this is the trend, this is radical, this is going to increase. So the cost could go from nominal to millions, literally, potentially. So, this is what customers are doing. Yeah, that's a great story. What makes sense on a financial, is there a cost of ownership? Is there a pattern for best practice for training? What do you guys advise cuz this is a lot of time and money involved in all potential, you know, good scenarios of upside. But you can get over your skis as they say, and be successful and be out of business if you don't manage it. I mean, that's what people are talking about, right? >> Yeah, absolutely. I think, you know, we see kind of three main vectors to reduce cost. I think one is make your deployment process easier overall, so that your engineering effort to even get your app running goes down. Two, would be get more from the compute you're already paying for, you're already paying, you know, for your instances in the cloud, but can you do more with that? And then three would be shop around for lower cost hardware to match your use case. So on the first one, I think making the deployment easier overall, there's a lot of manual work that goes into benchmarking, optimizing and packaging models for deployment. And because the performance of machine learning models can be really hardware dependent, you have to go through this process for each target you want to consider running your model on. And this is hard, you know, we see that every day. But for teams who want to incorporate some of these large language models into their applications, it might be desirable because licensing a model from a large vendor like OpenAI can leave you, you know, over provision, kind of paying for capabilities you don't need in your application or can lock you into them and you lose flexibility. So we have a customer whose team actually prepares models for deployment in a SaaS application that many of us use every day. And they told us recently that without kind of an automated benchmarking and experimentation platform, they were spending several days each to benchmark a single model on a single hardware type. So this is really, you know, manually intensive and then getting more from the compute you're already paying for. We do see customers who leave money on the table by running models that haven't been optimized specifically for the hardware target they're using, like Luis was mentioning. And for some teams they just don't have the time to go through an optimization process and for others they might lack kind of specialized expertise and this is something we can bring. And then on shopping around for different hardware types, we really see a huge variation in model performance across hardware, not just CPU vs. GPU, which is, you know, what people normally think of. But across CPU vendors themselves, high memory instances and across cloud providers even. So the best strategy here is for teams to really be able to, we say, look before you leap by running real world benchmarking and not just simulations or predictions to find the best software, hardware combination for their workload. >> Yeah. You guys sound like you have a very impressive customer base deploying large language models. Where would you categorize your current customer base? And as you look out, as you guys are growing, you have new customers coming in, take me through the progression. Take me through the profile of some of your customers you have now, size, are they hyperscalers, are they big app folks, are they kicking the tires? And then as people are out there scratching heads, I got to get in this game, what's their psychology like? Are they coming in with specific problems or do they have specific orientation point of view about what they want to do? Can you share some data around what you're seeing? >> Yeah, I think, you know, we have customers that kind of range across the spectrum of sophistication from teams that basically don't have MLOps expertise in their company at all. And so they're really looking for us to kind of give a full service, how should I do everything from, you know, optimization, find the hardware, prepare for deployment. And then we have teams that, you know, maybe already have their serving and hosting infrastructure up and ready and they already have models in production and they're really just looking to, you know, take the extra juice out of the hardware and just do really specific on that optimization piece. I think one place where we're doing a lot more work now is kind of in the developer tooling, you know, model selection space. And that's kind of an area that we're creating more tools for, particularly within the PyTorch ecosystem to bring kind of this power earlier in the development cycle so that as people are grabbing a model off the shelf, they can, you know, see how it might perform and use that to inform their development process. >> Luis, what's the big, I like this idea of picking the models because isn't that like going to the market and picking the best model for your data? It's like, you know, it's like, isn't there a certain approaches? What's your view on this? 'Cause this is where everyone, I think it's going to be a land rush for this and I want to get your thoughts. >> For sure, yeah. So, you know, I guess I'll start with saying the one main takeaway that we got from the GPT-J study is that, you know, having a different understanding of what your model's compute and memory requirements are, very quickly, early on helps with the much smarter AI model deployments, right? So, and in fact, you know, Anna just touched on this, but I want to, you know, make sure that it's clear that OctoML is putting that power into user's hands right now. So in partnership with AWS, we are launching this new PyTorch native profiler that allows you with a single, you know, one line, you know, code decorator allows you to see how your code runs on a variety of different hardware after accelerations. So it gives you very clear, you know, data on how you should think about your model deployments. And this ties back to choices of models. So like, if you have a set of choices that are equally good of models in terms of functionality and you want to understand after acceleration how are you going to deploy, how much they're going to cost or what are the options using a automated process of making a decision is really, really useful. And in fact, so I think these events can get early access to this by signing up for the Octopods, you know, this is exclusive group for insiders here, so you can go to OctoML.ai/pods to sign up. >> So that Octopod, is that a program? What is that, is that access to code? Is that a beta, what is that? Explain, take a minute and explain Octopod. >> I think the Octopod would be a group of people who is interested in experiencing this functionality. So it is the friends and users of OctoML that would be the Octopod. And then yes, after you sign up, we would provide you essentially the tool in code form for you to try out in your own. I mean, part of the benefit of this is that it happens in your own local environment and you're in control of everything kind of within the workflow that developers are already using to create and begin putting these models into their applications. So it would all be within your control. >> Got it. I think the big question I have for you is when do you, when does that one of your customers know they need to call you? What's their environment look like? What are they struggling with? What are the conversations they might be having on their side of the fence? If anyone's watching this, they're like, "Hey, you know what, I've got my team, we have a lot of data. Do we have our own language model or do I use someone else's?" There's a lot of this, I will say discovery going on around what to do, what path to take, what does that customer look like, if someone's listening, when do they know to call you guys, OctoML? >> Well, I mean the most obvious one is that you have a significant spend on AI/ML, come and talk to us, you know, putting AIML into production. So that's the clear one. In fact, just this morning I was talking to someone who is in life sciences space and is having, you know, 15 to $20 million a year cloud related to AI/ML deployment is a clear, it's a pretty clear match right there, right? So that's on the cost side. But I also want to emphasize something that Anna said earlier that, you know, the hardware and software complexity involved in putting model into production is really high. So we've been able to abstract that away, offering a clean automation flow enables one, to experiment early on, you know, how models would run and get them to production. And then two, once they are into production, gives you an automated flow to continuously updating your model and taking advantage of all this acceleration and ability to run the model on the right hardware. So anyways, let's say one then is cost, you know, you have significant cost and then two, you have an automation needs. And Anna please compliment that. >> Yeah, Anna you can please- >> Yeah, I think that's exactly right. Maybe the other time is when you are expecting a big scale up in serving your application, right? You're launching a new feature, you expect to get a lot of usage or, and you want to kind of anticipate maybe your CTO, your CIO, whoever pays your cloud bills is going to come after you, right? And so they want to know, you know, what's the return on putting this model essentially into my application stack? Am I going to, is the usage going to match what I'm paying for it? And then you can understand that. >> So you guys have a lot of the early adopters, they got big data teams, they're pushed in the production, they want to get a little QA, test the waters, understand, use your technology to figure it out. Is there any cases where people have gone into production, they have to pull it out? It's like the old lemon laws with your car, you buy a car and oh my god, it's not the way I wanted it. I mean, I can imagine the early people through the wall, so to speak, in the wave here are going to be bloody in the sense that they've gone in and tried stuff and get stuck with huge bills. Are you seeing that? Are people pulling stuff out of production and redeploying? Or I can imagine that if I had a bad deployment, I'd want to refactor that or actually replatform that. Do you see that too? >> Definitely after a sticker shock, yes, your customers will come and make sure that, you know, the sticker shock won't happen again. >> Yeah. >> But then there's another more thorough aspect here that I think we likely touched on, be worth elaborating a bit more is just how are you going to scale in a way that's feasible depending on the allocation that you get, right? So as we mentioned several times here, you know, model deployment is so hardware dependent and so complex that you tend to get a model for a hardware choice and then you want to scale that specific type of instance. But what if, when you want to scale because suddenly luckily got popular and, you know, you want to scale it up and then you don't have that instance anymore. So how do you live with whatever you have at that moment is something that we see customers needing as well. You know, so in fact, ideally what we want is customers to not think about what kind of specific instances they want. What they want is to know what their models need. Say, they know the SLA and then find a set of hybrid targets and instances that hit the SLA whenever they're also scaling, they're going to scale with more freedom, right? Instead of having to wait for AWS to give them more specific allocation for a specific instance. What if you could live with other types of hardware and scale up in a more free way, right? So that's another thing that we see customers, you know, like they need more freedom to be able to scale with whatever is available. >> Anna, you touched on this with the business model impact to that 6 million cost, if that goes out of control, there's a business model aspect and there's a technical operation aspect to the cost side too. You want to be mindful of riding the wave in a good way, but not getting over your skis. So that brings up the point around, you know, confidence, right? And teamwork. Because if you're in production, there's probably a team behind it. Talk about the team aspect of your customers. I mean, they're dedicated, they go put stuff into production, they're developers, there're data. What's in it for them? Are they getting better, are they in the beach, you know, reading the book. Are they, you know, are there easy street for them? What's the customer benefit to the teams? >> Yeah, absolutely. With just a few clicks of a button, you're in production, right? That's the dream. So yeah, I mean I think that, you know, we illustrated it before a little bit. I think the automated kind of benchmarking and optimization process, like when you think about the effort it takes to get that data by hand, which is what people are doing today, they just don't do it. So they're making decisions without the best information because it's, you know, there just isn't the bandwidth to get the information that they need to make the best decision and then know exactly how to deploy it. So I think it's actually bringing kind of a new insight and capability to these teams that they didn't have before. And then maybe another aspect on the team side is that it's making the hand-off of the models from the data science teams to the model deployment teams more seamless. So we have, you know, we have seen in the past that this kind of transition point is the place where there are a lot of hiccups, right? The data science team will give a model to the production team and it'll be too slow for the application or it'll be too expensive to run and it has to go back and be changed and kind of this loop. And so, you know, with the PyTorch profiler that Luis was talking about, and then also, you know, the other ways we do optimization that kind of prevents that hand-off problem from happening. >> Luis and Anna, you guys have a great company. Final couple minutes left. Talk about the company, the people there, what's the culture like, you know, if Intel has Moore's law, which is, you know, doubling the performance in few years, what's the culture like there? Is it, you know, more throughput, better pricing? Explain what's going on with the company and put a plug in. Luis, we'll start with you. >> Yeah, absolutely. I'm extremely proud of the team that we built here. You know, we have a people first culture, you know, very, very collaborative and folks, we all have a shared mission here of making AI more accessible and sustainable. We have a very diverse team in terms of backgrounds and life stories, you know, to do what we do here, we need a team that has expertise in software engineering, in machine learning, in computer architecture. Even though we don't build chips, we need to understand how they work, right? So, and then, you know, the fact that we have this, this very really, really varied set of backgrounds makes the environment, you know, it's say very exciting to learn more about, you know, assistance end-to-end. But also makes it for a very interesting, you know, work environment, right? So people have different backgrounds, different stories. Some of them went to grad school, others, you know, were in intelligence agencies and now are working here, you know. So we have a really interesting set of people and, you know, life is too short not to work with interesting humans. You know, that's something that I like to think about, you know. >> I'm sure your off-site meetings are a lot of fun, people talking about computer architectures, silicon advances, the next GPU, the big data models coming in. Anna, what's your take? What's the culture like? What's the company vibe and what are you guys looking to do? What's the customer success pattern? What's up? >> Yeah, absolutely. I mean, I, you know, second all of the great things that Luis just said about the team. I think one that I, an additional one that I'd really like to underscore is kind of this customer obsession, to use a term you all know well. And focus on the end users and really making the experiences that we're bringing to our user who are developers really, you know, useful and valuable for them. And so I think, you know, all of these tools that we're trying to put in the hands of users, the industry and the market is changing so rapidly that our products across the board, you know, all of the companies that, you know, are part of the showcase today, we're all evolving them so quickly and we can only do that kind of really hand in glove with our users. So that would be another thing I'd emphasize. >> I think the change dynamic, the power dynamics of this industry is just the beginning. I'm very bullish that this is going to be probably one of the biggest inflection points in history of the computer industry because of all the dynamics of the confluence of all the forces, which you mentioned some of them, I mean PC, you know, interoperability within internetworking and you got, you know, the web and then mobile. Now we have this, I mean, I wouldn't even put social media even in the close to this. Like, this is like, changes user experience, changes infrastructure. There's going to be massive accelerations in performance on the hardware side from AWS's of the world and cloud and you got the edge and more data. This is really what big data was going to look like. This is the beginning. Final question, what do you guys see going forward in the future? >> Well, it's undeniable that machine learning and AI models are becoming an integral part of an interesting application today, right? So, and the clear trends here are, you know, more and more competitional needs for these models because they're only getting more and more powerful. And then two, you know, seeing the complexity of the infrastructure where they run, you know, just considering the cloud, there's like a wide variety of choices there, right? So being able to live with that and making the most out of it in a way that does not require, you know, an impossible to find team is something that's pretty clear. So the need for automation, abstracting with the complexity is definitely here. And we are seeing this, you know, trends are that you also see models starting to move to the edge as well. So it's clear that we're seeing, we are going to live in a world where there's no large models living in the cloud. And then, you know, edge models that talk to these models in the cloud to form, you know, an end-to-end truly intelligent application. >> Anna? >> Yeah, I think, you know, our, Luis said it at the beginning. Our vision is to make AI sustainable and accessible. And I think as this technology just expands in every company and every team, that's going to happen kind of on its own. And we're here to help support that. And I think you can't do that without tools like those like OctoML. >> I think it's going to be an error of massive invention, creativity, a lot of the format heavy lifting is going to allow the talented people to automate their intellect. I mean, this is really kind of what we see going on. And Luis, thank you so much. Anna, thanks for coming on this segment. Thanks for coming on theCUBE and being part of the AWS Startup Showcase. I'm John Furrier, your host. Thanks for watching. (upbeat music)

(upbeat music) >> Hello everyone. Welcome to theCube's presentation of the "AWS Startup Showcase." The topic this episode is AI and machine learning, top startups building foundational model infrastructure. This is season three, episode one of the ongoing series covering exciting startups from the AWS ecosystem. And this time we're talking about AI and machine learning. I'm your host, John Furrier. I'm excited I'm joined today by Robert Nishihara, who's the co-founder and CEO of a hot startup called Anyscale. He's here to talk about Ray, the open source project, Anyscale's infrastructure for foundation as well. Robert, thank you for joining us today. >> Yeah, thanks so much as well. >> I've been following your company since the founding pre pandemic and you guys really had a great vision scaled up and in a perfect position for this big wave that we all see with ChatGPT and OpenAI that's gone mainstream. Finally, AI has broken out through the ropes and now gone mainstream, so I think you guys are really well positioned. I'm looking forward to to talking with you today. But before we get into it, introduce the core mission for Anyscale. Why do you guys exist? What is the North Star for Anyscale? >> Yeah, like you mentioned, there's a tremendous amount of excitement about AI right now. You know, I think a lot of us believe that AI can transform just every different industry. So one of the things that was clear to us when we started this company was that the amount of compute needed to do AI was just exploding. Like to actually succeed with AI, companies like OpenAI or Google or you know, these companies getting a lot of value from AI, were not just running these machine learning models on their laptops or on a single machine. They were scaling these applications across hundreds or thousands or more machines and GPUs and other resources in the Cloud. And so to actually succeed with AI, and this has been one of the biggest trends in computing, maybe the biggest trend in computing in, you know, in recent history, the amount of compute has been exploding. And so to actually succeed with that AI, to actually build these scalable applications and scale the AI applications, there's a tremendous software engineering lift to build the infrastructure to actually run these scalable applications. And that's very hard to do. So one of the reasons many AI projects and initiatives fail is that, or don't make it to production, is the need for this scale, the infrastructure lift, to actually make it happen. So our goal here with Anyscale and Ray, is to make that easy, is to make scalable computing easy. So that as a developer or as a business, if you want to do AI, if you want to get value out of AI, all you need to know is how to program on your laptop. Like, all you need to know is how to program in Python. And if you can do that, then you're good to go. Then you can do what companies like OpenAI or Google do and get value out of machine learning. >> That programming example of how easy it is with Python reminds me of the early days of Cloud, when infrastructure as code was talked about was, it was just code the infrastructure programmable. That's super important. That's what AI people wanted, first program AI. That's the new trend. And I want to understand, if you don't mind explaining, the relationship that Anyscale has to these foundational models and particular the large language models, also called LLMs, was seen with like OpenAI and ChatGPT. Before you get into the relationship that you have with them, can you explain why the hype around foundational models? Why are people going crazy over foundational models? What is it and why is it so important? >> Yeah, so foundational models and foundation models are incredibly important because they enable businesses and developers to get value out of machine learning, to use machine learning off the shelf with these large models that have been trained on tons of data and that are useful out of the box. And then, of course, you know, as a business or as a developer, you can take those foundational models and repurpose them or fine tune them or adapt them to your specific use case and what you want to achieve. But it's much easier to do that than to train them from scratch. And I think there are three, for people to actually use foundation models, there are three main types of workloads or problems that need to be solved. One is training these foundation models in the first place, like actually creating them. The second is fine tuning them and adapting them to your use case. And the third is serving them and actually deploying them. Okay, so Ray and Anyscale are used for all of these three different workloads. Companies like OpenAI or Cohere that train large language models. Or open source versions like GPTJ are done on top of Ray. There are many startups and other businesses that fine tune, that, you know, don't want to train the large underlying foundation models, but that do want to fine tune them, do want to adapt them to their purposes, and build products around them and serve them, those are also using Ray and Anyscale for that fine tuning and that serving. And so the reason that Ray and Anyscale are important here is that, you know, building and using foundation models requires a huge scale. It requires a lot of data. It requires a lot of compute, GPUs, TPUs, other resources. And to actually take advantage of that and actually build these scalable applications, there's a lot of infrastructure that needs to happen under the hood. And so you can either use Ray and Anyscale to take care of that and manage the infrastructure and solve those infrastructure problems. Or you can build the infrastructure and manage the infrastructure yourself, which you can do, but it's going to slow your team down. It's going to, you know, many of the businesses we work with simply don't want to be in the business of managing infrastructure and building infrastructure. They want to focus on product development and move faster. >> I know you got a keynote presentation we're going to go to in a second, but I think you hit on something I think is the real tipping point, doing it yourself, hard to do. These are things where opportunities are and the Cloud did that with data centers. Turned a data center and made it an API. The heavy lifting went away and went to the Cloud so people could be more creative and build their product. In this case, build their creativity. Is that kind of what's the big deal? Is that kind of a big deal happening that you guys are taking the learnings and making that available so people don't have to do that? >> That's exactly right. So today, if you want to succeed with AI, if you want to use AI in your business, infrastructure work is on the critical path for doing that. To do AI, you have to build infrastructure. You have to figure out how to scale your applications. That's going to change. We're going to get to the point, and you know, with Ray and Anyscale, we're going to remove the infrastructure from the critical path so that as a developer or as a business, all you need to focus on is your application logic, what you want the the program to do, what you want your application to do, how you want the AI to actually interface with the rest of your product. Now the way that will happen is that Ray and Anyscale will still, the infrastructure work will still happen. It'll just be under the hood and taken care of by Ray in Anyscale. And so I think something like this is really necessary for AI to reach its potential, for AI to have the impact and the reach that we think it will, you have to make it easier to do. >> And just for clarification to point out, if you don't mind explaining the relationship of Ray and Anyscale real quick just before we get into the presentation. >> So Ray is an open source project. We created it. We were at Berkeley doing machine learning. We started Ray so that, in order to provide an easy, a simple open source tool for building and running scalable applications. And Anyscale is the managed version of Ray, basically we will run Ray for you in the Cloud, provide a lot of tools around the developer experience and managing the infrastructure and providing more performance and superior infrastructure. >> Awesome. I know you got a presentation on Ray and Anyscale and you guys are positioning as the infrastructure for foundational models. So I'll let you take it away and then when you're done presenting, we'll come back, I'll probably grill you with a few questions and then we'll close it out so take it away. >> Robert: Sounds great. So I'll say a little bit about how companies are using Ray and Anyscale for foundation models. The first thing I want to mention is just why we're doing this in the first place. And the underlying observation, the underlying trend here, and this is a plot from OpenAI, is that the amount of compute needed to do machine learning has been exploding. It's been growing at something like 35 times every 18 months. This is absolutely enormous. And other people have written papers measuring this trend and you get different numbers. But the point is, no matter how you slice and dice it, it' a astronomical rate. Now if you compare that to something we're all familiar with, like Moore's Law, which says that, you know, the processor performance doubles every roughly 18 months, you can see that there's just a tremendous gap between the needs, the compute needs of machine learning applications, and what you can do with a single chip, right. So even if Moore's Law were continuing strong and you know, doing what it used to be doing, even if that were the case, there would still be a tremendous gap between what you can do with the chip and what you need in order to do machine learning. And so given this graph, what we've seen, and what has been clear to us since we started this company, is that doing AI requires scaling. There's no way around it. It's not a nice to have, it's really a requirement. And so that led us to start Ray, which is the open source project that we started to make it easy to build these scalable Python applications and scalable machine learning applications. And since we started the project, it's been adopted by a tremendous number of companies. Companies like OpenAI, which use Ray to train their large models like ChatGPT, companies like Uber, which run all of their deep learning and classical machine learning on top of Ray, companies like Shopify or Spotify or Instacart or Lyft or Netflix, ByteDance, which use Ray for their machine learning infrastructure. Companies like Ant Group, which makes Alipay, you know, they use Ray across the board for fraud detection, for online learning, for detecting money laundering, you know, for graph processing, stream processing. Companies like Amazon, you know, run Ray at a tremendous scale and just petabytes of data every single day. And so the project has seen just enormous adoption since, over the past few years. And one of the most exciting use cases is really providing the infrastructure for building training, fine tuning, and serving foundation models. So I'll say a little bit about, you know, here are some examples of companies using Ray for foundation models. Cohere trains large language models. OpenAI also trains large language models. You can think about the workloads required there are things like supervised pre-training, also reinforcement learning from human feedback. So this is not only the regular supervised learning, but actually more complex reinforcement learning workloads that take human input about what response to a particular question, you know is better than a certain other response. And incorporating that into the learning. There's open source versions as well, like GPTJ also built on top of Ray as well as projects like Alpa coming out of UC Berkeley. So these are some of the examples of exciting projects in organizations, training and creating these large language models and serving them using Ray. Okay, so what actually is Ray? Well, there are two layers to Ray. At the lowest level, there's the core Ray system. This is essentially low level primitives for building scalable Python applications. Things like taking a Python function or a Python class and executing them in the cluster setting. So Ray core is extremely flexible and you can build arbitrary scalable applications on top of Ray. So on top of Ray, on top of the core system, what really gives Ray a lot of its power is this ecosystem of scalable libraries. So on top of the core system you have libraries, scalable libraries for ingesting and pre-processing data, for training your models, for fine tuning those models, for hyper parameter tuning, for doing batch processing and batch inference, for doing model serving and deployment, right. And a lot of the Ray users, the reason they like Ray is that they want to run multiple workloads. They want to train and serve their models, right. They want to load their data and feed that into training. And Ray provides common infrastructure for all of these different workloads. So this is a little overview of what Ray, the different components of Ray. So why do people choose to go with Ray? I think there are three main reasons. The first is the unified nature. The fact that it is common infrastructure for scaling arbitrary workloads, from data ingest to pre-processing to training to inference and serving, right. This also includes the fact that it's future proof. AI is incredibly fast moving. And so many people, many companies that have built their own machine learning infrastructure and standardized on particular workflows for doing machine learning have found that their workflows are too rigid to enable new capabilities. If they want to do reinforcement learning, if they want to use graph neural networks, they don't have a way of doing that with their standard tooling. And so Ray, being future proof and being flexible and general gives them that ability. Another reason people choose Ray in Anyscale is the scalability. This is really our bread and butter. This is the reason, the whole point of Ray, you know, making it easy to go from your laptop to running on thousands of GPUs, making it easy to scale your development workloads and run them in production, making it easy to scale, you know, training to scale data ingest, pre-processing and so on. So scalability and performance, you know, are critical for doing machine learning and that is something that Ray provides out of the box. And lastly, Ray is an open ecosystem. You can run it anywhere. You can run it on any Cloud provider. Google, you know, Google Cloud, AWS, Asure. You can run it on your Kubernetes cluster. You can run it on your laptop. It's extremely portable. And not only that, it's framework agnostic. You can use Ray to scale arbitrary Python workloads. You can use it to scale and it integrates with libraries like TensorFlow or PyTorch or JAX or XG Boost or Hugging Face or PyTorch Lightning, right, or Scikit-learn or just your own arbitrary Python code. It's open source. And in addition to integrating with the rest of the machine learning ecosystem and these machine learning frameworks, you can use Ray along with all of the other tooling in the machine learning ecosystem. That's things like weights and biases or ML flow, right. Or you know, different data platforms like Databricks, you know, Delta Lake or Snowflake or tools for model monitoring for feature stores, all of these integrate with Ray. And that's, you know, Ray provides that kind of flexibility so that you can integrate it into the rest of your workflow. And then Anyscale is the scalable compute platform that's built on top, you know, that provides Ray. So Anyscale is a managed Ray service that runs in the Cloud. And what Anyscale does is it offers the best way to run Ray. And if you think about what you get with Anyscale, there are fundamentally two things. One is about moving faster, accelerating the time to market. And you get that by having the managed service so that as a developer you don't have to worry about managing infrastructure, you don't have to worry about configuring infrastructure. You also, it provides, you know, optimized developer workflows. Things like easily moving from development to production, things like having the observability tooling, the debug ability to actually easily diagnose what's going wrong in a distributed application. So things like the dashboards and the other other kinds of tooling for collaboration, for monitoring and so on. And then on top of that, so that's the first bucket, developer productivity, moving faster, faster experimentation and iteration. The second reason that people choose Anyscale is superior infrastructure. So this is things like, you know, cost deficiency, being able to easily take advantage of spot instances, being able to get higher GPU utilization, things like faster cluster startup times and auto scaling. Things like just overall better performance and faster scheduling. And so these are the kinds of things that Anyscale provides on top of Ray. It's the managed infrastructure. It's fast, it's like the developer productivity and velocity as well as performance. So this is what I wanted to share about Ray in Anyscale. >> John: Awesome. >> Provide that context. But John, I'm curious what you think. >> I love it. I love the, so first of all, it's a platform because that's the platform architecture right there. So just to clarify, this is an Anyscale platform, not- >> That's right. >> Tools. So you got tools in the platform. Okay, that's key. Love that managed service. Just curious, you mentioned Python multiple times, is that because of PyTorch and TensorFlow or Python's the most friendly with machine learning or it's because it's very common amongst all developers? >> That's a great question. Python is the language that people are using to do machine learning. So it's the natural starting point. Now, of course, Ray is actually designed in a language agnostic way and there are companies out there that use Ray to build scalable Java applications. But for the most part right now we're focused on Python and being the best way to build these scalable Python and machine learning applications. But, of course, down the road there always is that potential. >> So if you're slinging Python code out there and you're watching that, you're watching this video, get on Anyscale bus quickly. Also, I just, while you were giving the presentation, I couldn't help, since you mentioned OpenAI, which by the way, congratulations 'cause they've had great scale, I've noticed in their rapid growth 'cause they were the fastest company to the number of users than anyone in the history of the computer industry, so major successor, OpenAI and ChatGPT, huge fan. I'm not a skeptic at all. I think it's just the beginning, so congratulations. But I actually typed into ChatGPT, what are the top three benefits of Anyscale and came up with scalability, flexibility, and ease of use. Obviously, scalability is what you guys are called. >> That's pretty good. >> So that's what they came up with. So they nailed it. Did you have an inside prompt training, buy it there? Only kidding. (Robert laughs) >> Yeah, we hard coded that one. >> But that's the kind of thing that came up really, really quickly if I asked it to write a sales document, it probably will, but this is the future interface. This is why people are getting excited about the foundational models and the large language models because it's allowing the interface with the user, the consumer, to be more human, more natural. And this is clearly will be in every application in the future. >> Absolutely. This is how people are going to interface with software, how they're going to interface with products in the future. It's not just something, you know, not just a chat bot that you talk to. This is going to be how you get things done, right. How you use your web browser or how you use, you know, how you use Photoshop or how you use other products. Like you're not going to spend hours learning all the APIs and how to use them. You're going to talk to it and tell it what you want it to do. And of course, you know, if it doesn't understand it, it's going to ask clarifying questions. You're going to have a conversation and then it'll figure it out. >> This is going to be one of those things, we're going to look back at this time Robert and saying, "Yeah, from that company, that was the beginning of that wave." And just like AWS and Cloud Computing, the folks who got in early really were in position when say the pandemic came. So getting in early is a good thing and that's what everyone's talking about is getting in early and playing around, maybe replatforming or even picking one or few apps to refactor with some staff and managed services. So people are definitely jumping in. So I have to ask you the ROI cost question. You mentioned some of those, Moore's Law versus what's going on in the industry. When you look at that kind of scale, the first thing that jumps out at people is, "Okay, I love it. Let's go play around." But what's it going to cost me? Am I going to be tied to certain GPUs? What's the landscape look like from an operational standpoint, from the customer? Are they locked in and the benefit was flexibility, are you flexible to handle any Cloud? What is the customers, what are they looking at? Basically, that's my question. What's the customer looking at? >> Cost is super important here and many of the companies, I mean, companies are spending a huge amount on their Cloud computing, on AWS, and on doing AI, right. And I think a lot of the advantage of Anyscale, what we can provide here is not only better performance, but cost efficiency. Because if we can run something faster and more efficiently, it can also use less resources and you can lower your Cloud spending, right. We've seen companies go from, you know, 20% GPU utilization with their current setup and the current tools they're using to running on Anyscale and getting more like 95, you know, 100% GPU utilization. That's something like a five x improvement right there. So depending on the kind of application you're running, you know, it's a significant cost savings. We've seen companies that have, you know, processing petabytes of data every single day with Ray going from, you know, getting order of magnitude cost savings by switching from what they were previously doing to running their application on Ray. And when you have applications that are spending, you know, potentially $100 million a year and getting a 10 X cost savings is just absolutely enormous. So these are some of the kinds of- >> Data infrastructure is super important. Again, if the customer, if you're a prospect to this and thinking about going in here, just like the Cloud, you got infrastructure, you got the platform, you got SaaS, same kind of thing's going to go on in AI. So I want to get into that, you know, ROI discussion and some of the impact with your customers that are leveraging the platform. But first I hear you got a demo. >> Robert: Yeah, so let me show you, let me give you a quick run through here. So what I have open here is the Anyscale UI. I've started a little Anyscale Workspace. So Workspaces are the Anyscale concept for interactive developments, right. So here, imagine I'm just, you want to have a familiar experience like you're developing on your laptop. And here I have a terminal. It's not on my laptop. It's actually in the cloud running on Anyscale. And I'm just going to kick this off. This is going to train a large language model, so OPT. And it's doing this on 32 GPUs. We've got a cluster here with a bunch of CPU cores, bunch of memory. And as that's running, and by the way, if I wanted to run this on instead of 32 GPUs, 64, 128, this is just a one line change when I launch the Workspace. And what I can do is I can pull up VS code, right. Remember this is the interactive development experience. I can look at the actual code. Here it's using Ray train to train the torch model. We've got the training loop and we're saying that each worker gets access to one GPU and four CPU cores. And, of course, as I make the model larger, this is using deep speed, as I make the model larger, I could increase the number of GPUs that each worker gets access to, right. And how that is distributed across the cluster. And if I wanted to run on CPUs instead of GPUs or a different, you know, accelerator type, again, this is just a one line change. And here we're using Ray train to train the models, just taking my vanilla PyTorch model using Hugging Face and then scaling that across a bunch of GPUs. And, of course, if I want to look at the dashboard, I can go to the Ray dashboard. There are a bunch of different visualizations I can look at. I can look at the GPU utilization. I can look at, you know, the CPU utilization here where I think we're currently loading the model and running that actual application to start the training. And some of the things that are really convenient here about Anyscale, both I can get that interactive development experience with VS code. You know, I can look at the dashboards. I can monitor what's going on. It feels, I have a terminal, it feels like my laptop, but it's actually running on a large cluster. And I can, with however many GPUs or other resources that I want. And so it's really trying to combine the best of having the familiar experience of programming on your laptop, but with the benefits, you know, being able to take advantage of all the resources in the Cloud to scale. And it's like when, you know, you're talking about cost efficiency. One of the biggest reasons that people waste money, one of the silly reasons for wasting money is just forgetting to turn off your GPUs. And what you can do here is, of course, things will auto terminate if they're idle. But imagine you go to sleep, I have this big cluster. You can turn it off, shut off the cluster, come back tomorrow, restart the Workspace, and you know, your big cluster is back up and all of your code changes are still there. All of your local file edits. It's like you just closed your laptop and came back and opened it up again. And so this is the kind of experience we want to provide for our users. So that's what I wanted to share with you. >> Well, I think that whole, couple of things, lines of code change, single line of code change, that's game changing. And then the cost thing, I mean human error is a big deal. People pass out at their computer. They've been coding all night or they just forget about it. I mean, and then it's just like leaving the lights on or your water running in your house. It's just, at the scale that it is, the numbers will add up. That's a huge deal. So I think, you know, compute back in the old days, there's no compute. Okay, it's just compute sitting there idle. But you know, data cranking the models is doing, that's a big point. >> Another thing I want to add there about cost efficiency is that we make it really easy to use, if you're running on Anyscale, to use spot instances and these preemptable instances that can just be significantly cheaper than the on-demand instances. And so when we see our customers go from what they're doing before to using Anyscale and they go from not using these spot instances 'cause they don't have the infrastructure around it, the fault tolerance to handle the preemption and things like that, to being able to just check a box and use spot instances and save a bunch of money. >> You know, this was my whole, my feature article at Reinvent last year when I met with Adam Selipsky, this next gen Cloud is here. I mean, it's not auto scale, it's infrastructure scale. It's agility. It's flexibility. I think this is where the world needs to go. Almost what DevOps did for Cloud and what you were showing me that demo had this whole SRE vibe. And remember Google had site reliability engines to manage all those servers. This is kind of like an SRE vibe for data at scale. I mean, a similar kind of order of magnitude. I mean, I might be a little bit off base there, but how would you explain it? >> It's a nice analogy. I mean, what we are trying to do here is get to the point where developers don't think about infrastructure. Where developers only think about their application logic. And where businesses can do AI, can succeed with AI, and build these scalable applications, but they don't have to build, you know, an infrastructure team. They don't have to develop that expertise. They don't have to invest years in building their internal machine learning infrastructure. They can just focus on the Python code, on their application logic, and run the stuff out of the box. >> Awesome. Well, I appreciate the time. Before we wrap up here, give a plug for the company. I know you got a couple websites. Again, go, Ray's got its own website. You got Anyscale. You got an event coming up. Give a plug for the company looking to hire. Put a plug in for the company. >> Yeah, absolutely. Thank you. So first of all, you know, we think AI is really going to transform every industry and the opportunity is there, right. We can be the infrastructure that enables all of that to happen, that makes it easy for companies to succeed with AI, and get value out of AI. Now we have, if you're interested in learning more about Ray, Ray has been emerging as the standard way to build scalable applications. Our adoption has been exploding. I mentioned companies like OpenAI using Ray to train their models. But really across the board companies like Netflix and Cruise and Instacart and Lyft and Uber, you know, just among tech companies. It's across every industry. You know, gaming companies, agriculture, you know, farming, robotics, drug discovery, you know, FinTech, we see it across the board. And all of these companies can get value out of AI, can really use AI to improve their businesses. So if you're interested in learning more about Ray and Anyscale, we have our Ray Summit coming up in September. This is going to highlight a lot of the most impressive use cases and stories across the industry. And if your business, if you want to use LLMs, you want to train these LLMs, these large language models, you want to fine tune them with your data, you want to deploy them, serve them, and build applications and products around them, give us a call, talk to us. You know, we can really take the infrastructure piece, you know, off the critical path and make that easy for you. So that's what I would say. And, you know, like you mentioned, we're hiring across the board, you know, engineering, product, go-to-market, and it's an exciting time. >> Robert Nishihara, co-founder and CEO of Anyscale, congratulations on a great company you've built and continuing to iterate on and you got growth ahead of you, you got a tailwind. I mean, the AI wave is here. I think OpenAI and ChatGPT, a customer of yours, have really opened up the mainstream visibility into this new generation of applications, user interface, roll of data, large scale, how to make that programmable so we're going to need that infrastructure. So thanks for coming on this season three, episode one of the ongoing series of the hot startups. In this case, this episode is the top startups building foundational model infrastructure for AI and ML. I'm John Furrier, your host. Thanks for watching. (upbeat music)

(gentle music) >> Hello, and welcome to theCUBE's coverage of International Women's Day. I'm John Furrier, host of theCUBE. I've got a great guest here, CUBE alumni and very impressive, inspiring, Rachel Mushahwar Skaff, who's a managing director and general manager at AWS. Rachel, great to see you. Thanks for coming on. >> Thank you so much. It's always a pleasure to be here. You all make such a tremendous impact with reporting out what's happening in the tech space, and frankly, investing in topics like this, so thank you. >> It's our pleasure. Your career has been really impressive. You worked at Intel for almost a decade, and that company is very tech, very focused on Moore's law, cadence of technology power in the industry. Now at AWS, powering next-generation cloud. What inspired you to get into tech? How did you get here and how have you approached your career journey, because it's quite a track record? >> Wow, how long do we have? (Rachel and John laugh) >> John: We can go as long as you want. (laughs) It's great. >> You know, all joking aside, I think at the end of the day, it's about this simple statement. If you don't get goosebumps every single morning that you're waking up to do your job, it's not good enough. And that's a bit about how I've made all of the different career transitions that I have. You know, everything from building out data centers around the world, to leading network and engineering teams, to leading applications teams, to going and working for, you know, the largest semiconductor in the world, and now at AWS, every single one of those opportunities gave me goosebumps. And I was really focused on how do I surround myself with humans that are better than I am, smarter than I am, companies that plan in decades, but live in moments, companies that invest in their employees and create like artists? And frankly, for me, being part of a company where people know that life is finite, but they want to make an infinite impact, that's a bit about my career journey in a nutshell. >> Yeah. What's interesting is that, you know, over the years, a lot's changed, and a theme that we're hearing from leaders now that are heading up large teams and running companies, they have, you know, they have 20-plus years of experience under their belt and they look back and they say, "Wow, "things have changed and it's changing faster now, "hopefully faster to get change." But they all talk about confidence and they talk about curiosity and building. When did you know that this was going to be something that you got the goosebumps? And were there blockers in your way and how did you handle that? (Rachel laughs) >> There's always blockers in our way, and I think a lot of people don't actually talk about the blockers. I think they make it sound like, hey, I had this plan from day one, and every decision I've made has been perfect. And for me, I'll tell you, right, there are moments in your life that mark a differentiation and those moments that you realize nothing will be the same. And time is kind of divided into two parts, right, before this moment and after this moment. And that's everything from, before I had kids, that's a pretty big moment in people's lives, to after I had kids, and how do you work through some of those opportunities? Before I got married, before I got divorced. Before I went to this company, after I left this company. And I think the key for all of those is just having an insatiable curiosity around how do you continue to do better, create better and make better? And I'll tell you, those blockers, they exist. Coming back from maternity leave, hard. Coming back from a medical leave, hard. Coming back from caring for a sick parent or a sick friend, hard. But all of those things start to help craft who you are as a human being, not as a leader, but as a human being, and allows you to have some empathy with the people that you surround yourself with, right? And for me, it's, (sighs) you can think about these blockers in one of two ways. You can think about it as, you know, every single time that you're tempted to react in the same way to a blocker, you can be a prisoner of your past, or you can change how you react and be a pioneer of the future. It's not a blocker when you think about it in those terms. >> Mindset matters, and that's really a great point. You brought up something that's interesting, I want to bring this up. Some of the challenges in different stages of our lives. You know, one thing that's come out of this set of interviews, this, of day and in conversations is, that I haven't heard before, is the result of COVID, working at home brought empathy about people's personal lives to the table. That came up in a couple interviews. What's your reaction to that? Because that highlights that we're human, to your point of view. >> It does. It does. And I'm so thankful that you don't ask about balance because that is a pet peeve of mine, because there is no such thing as balance. If you're in perfect balance, you are not moving and you're not changing. But when you think about, you know, the impact of COVID and how the world has changed since that, it has allowed all of us to really think about, you know, what do we want to do versus what do we have to do? And I think so many times, in both our professional lives and our personal lives, we get caught up in doing what we think we have to do to get ahead versus taking a step back and saying, "Hey, what do I want to do? "And how do I become a, you know, "a better human?" And many times, John, I'm asked, "Hey, "how do you define success or achievement?" And, you know, my answer is really, for me, the greatest results that I've achieved, both personally and professionally, is when I eliminate the word success and balance from my vocabulary, and replace them with two words: What's my contribution and what's my impact? Those things make a difference, regardless of gender. And I'll tell you, none of it is easy, ever. I think all of us have been broken, we've been stretched, we've been burnt out. But I also think what we have to talk about as leaders in the industry is how we've also found endurance and resilience. And when we felt unsteady, we've continued to go forward, right? When we can't decide, the best answer is do what's uncomfortable. And all of those things really stemmed from a part of what happened with COVID. >> Yeah, yeah, I love the uncomfortable and the balance highlight. You mentioned being off balance. That means you're growing, you're not standing still. I want to get your thoughts on this because one thing that has come out again this year, and last year as well, is having a team with you when you do it. So if you're off balance and you're going to stretch, if you have a good team with you, that's where people help each other. Not just pick them up, but like maybe get 'em back on track again. So, but if you're solo, you fall, (laughs) you fall harder. So what's your reaction to that? 'Cause this has come up, and this comes up in team building, workforce formation, goal setting, contribution. What's your reaction to that? >> So my reaction to that that is pretty simple. Nobody gets there on their own at all, right? Passion and ambition can only take you so far. You've got to have people and teams that are supporting you. And here's the funny thing about people, and frankly, about being a leader that I think is really important: People don't follow for you. People follow for who you help them become. Think about that for a second. And when you think about all the amazing things that companies and teams are able to do, it's because of those people. And it's because you have leaders that are out there, inspiring them to take what they believe is impossible and turn it into the possible. That's the power of teams. >> Can you give an example of your approach on how you do that? How do you build your teams? How do you grow them? How do you lead them effectively and also make 'em inclusive, diverse and equitable? >> Whew. I'll give you a great example of some work that we're doing at AWS. This year at re:Invent, for the first time in its history, we've launched an initiative with theCUBE called Women of the Cloud. And part of Women of the Cloud is highlighting the business impact that so many of our partners, our customers and our employees have had on the social, on the economic and on the financials of many companies. They just haven't had the opportunity to tell their story. And at Amazon, right, it is absolutely integral to us to highlight those examples and continue to extend that ethos to our partners and our customers. And I think one of the things that I shared with you at re:Invent was, you know, as U2's Bono put it, (John laughs) "We'll build it better than we did before "and we are the people "that we've been waiting for." So if we're not out there, advocating and highlighting all the amazing things that other women are doing in the ecosystem, who will? >> Well, I've got to say, I want to give you props for that program. Not only was it groundbreaking, it's still running strong. And I saw some things on LinkedIn that were really impressive in its network effect. And I met at least half a dozen new people I never would have met before through some of that content interaction and engagement. And this is like the power of the current world. I mean, getting the voices out there creates momentum. And it's good for Amazon. It's not just personal brand building for my next job or whatever, you know, reason. It's sharing and it's attracting others, and it's causing people to connect and meet each other in that world. So it's still going strong. (laughs) And this program we did last year was part of Rachel Thornton, who's now at MessageBird, and Mary Camarata. They were the sponsors for this International Women's Day. They're not there anymore, so we decided we're going to do it again because the impact is so significant. We had the Amazon Education group on. It's amazing and it's free, and we've got to get the word out. I mean, talk about leveling up fast. You get in and you get trained and get certified, and there's a zillion jobs out (laughs) there in cloud, right, and partners. So this kind of leadership is really important. What was the key learnings that you've taken away and how do you extend this opportunity to nurture the talent out there in the field? Because when you throw the content out there from great leaders and practitioners and developers, it attracts other people. >> It does. It does. So look, I think there's two types of people, people that are focused on being and people who are focused on doing. And let me give you an example, right? When we think about labels of, hey, Rachel's a female executive who launched Women of the Cloud, that label really limits me. I'd rather just be a great executive. Or, hey, there's a great entrepreneur. Let's not be a great entrepreneur. Just go build something and sell it. And that's part of this whole Women of the cloud, is I don't want people focused on what their label is. I want people sharing their stories about what they're doing, and that's where the lasting impact happens, right? I think about something that my grandmother used to tell me, and she used to tell me, "Rachel, how successful "you are, doesn't matter. "The lasting impact that you have "is your legacy in this very finite time "that you have on Earth. "Leave a legacy." And that's what Women of the Cloud is about. So that people can start to say, "Oh, geez, "I didn't know that that was possible. "I didn't think about my career in that way." And, you know, all of those different types of stories that you're hearing out there. >> And I want to highlight something you said. We had another Amazonian on the program for this day earlier and she coined a term, 'cause inside Amazon, you have common language. One of them is bar raising. Raise the bar, that's an Amazonian (Rachel laughs) term. It means contribute and improve and raise the bar of capability. She said, "Bar raising is gender neutral. "The bar is a bar." And I'm like, wow, that was amazing. Now, that means your contribution angle there highlights that. What's the biggest challenge to get that mindset set in culture, in these- >> Oh. >> 'Cause it's that simple, contribution is neutral. >> It absolutely is neutral, but it's like I said earlier, I think so many times, people are focused on success and being a great leader versus what's the contribution I'm making and how am I doing as a leader, you know? And when it comes to a lot of the leadership principles that Amazon has, including bar raising, which means insisting on the highest standards, and then those standards continue to raise every single time. And what that is all about is having all of our employees figure out, how do I get better every single day, right? That's what it's about. It's not about being better than the peer next to you. It's about how do I become a better leader, a better human being than I was yesterday? >> Awesome. >> You know, I read this really cute quote and I think it really resonates. "You meditate to upgrade your software "and you work out to upgrade your hardware." And while it's important that we're all ourselves at work, we can't deny that a lot of times, ourselves still need that meditation or that workout. >> Well, I hope I don't have any zero days in my software out there, so, but I'm going to definitely work on that. I love that quote. I'm going to use that. Thank you very much. That was awesome. I got to ask you, I know you're really passionate about, and we've talked about this, around, so you're a great leader but you're also focused on what's behind you in the generation, pipelining women leaders, okay? Seats at the table, mentoring and sponsorship. What can we do to build a strong pipeline of leaders in technology and business? And where do you see the biggest opportunity to nurture the talent in these fields? >> Hmm, you know, that's great, great question. And, you know, I just read a "Forbes" article by another Amazonian, Tanuja Randery, who talked about, you know, some really interesting stats. And one of the stats that she shared was, you know, by 2030, less than 25% of tech specialists will be female, less than 25%. That's only a 6% growth from where we are in 2023, so in seven years. That's alarming. So we've really got to figure out what are the kinds of things that we're going to go do from an Amazon perspective to impact that? And one of the obvious starting points is showcasing tech careers to girls and young women, and talking openly about what a technology career looks like. So specifically at Amazon, we've got an AWS Git IT program that helps schools and educators bring in tech role models to show them what potential careers look like in tech. I think that's one great way that we can help build the pipeline, but once we get the pipeline, we also have to figure out how we don't let that pipeline leak. Meaning how do we keep women and, you know, young women on their tech career? And I think big part of that, John, is really talking about how hard it is, but it's also greater than you can ever imagine. And letting them see executives that are very authentic and will talk about, geez, you know, the challenges of COVID were a time of crisis and accelerated change, and here's what it meant to me personally and here's what we were able to solve professionally. These younger generations are all about social impact, they're about economic impact and they're about financial impact. And if we're not talking about all three of those, both from how AWS is leading from the front, but how its executives are also taking that into their personal lives, they're not going to want to go into tech. >> Yeah, and I think one of the things you mentioned there about getting people that get IT, good call out there, but also, Amazon's going to train 30 million people, put hundreds of millions of dollars into education. And not only are they making it easier to get in to get trained, but once you're in, even savvy folks that are in there still have to accelerate. And there's more ways to level up, more things are happening, but there's a big trend around people changing careers either in their late 20s, early 30s, or even those moments you talk about, where it's before and after, even later in the careers, 40s, 50s. Leaders like, well, good experience, good training, who were in another discipline who re-skilled. So you have, you know, more certifications coming in. So there's still other pivot points in the pipeline. It's not just down here. And that, I find that interesting. Are you seeing that same leadership opportunities coming in where someone can come into tech older? >> Absolutely. You know, we've got some amazing programs, like Amazon Returnity, that really focuses on how do we get other, you know, how do we get women that have taken some time off of work to get back into the workforce? And here's the other thing about switching careers. If I look back on my career, I started out as a civil engineer, heavy highway construction. And now I lead a sales team at the largest cloud company in the world. And there were, you know, twists and turns around there. I've always focused on how do we change and how do we continue to evolve? So it's not just focused on, you know, young women in the pipeline. It's focused on all gender and all diverse types throughout their career, and making sure that we're providing an inclusive environment for them to bring in their unique skillsets. >> Yeah, a building has good steel. It's well structured. Roads have great foundations. You know, you got the builder in you there. >> Yes. >> So I have to ask you, what's on your mind as a tech athlete, as an executive at AWS? You know, you got your huge team, big goals, the economy's got a little bit of a headwind, but still, cloud's transforming, edge is exploding. What's your outlook as you look out in the tech landscape these days and how are you thinking about it? What your plans? Can you share a little bit about what's on your mind? >> Sure. So, geez, there's so many trends that are top of mind right now. Everything from zero trust to artificial intelligence to security. We have more access to data now than ever before. So the opportunities are limitless when we think about how we can apply technology to solve some really difficult customer problems, right? Innovation sometimes feels like it's happening at a rapid pace. And I also say, you know, there are years when nothing happens, and then there's years when centuries happen. And I feel like we're kind of in those years where centuries are happening. Cloud technologies are refining sports as we know them now. There's a surge of innovation in smart energy. Everyone's supply chain is looking to transform. Custom silicon is going mainstream. And frankly, AWS's customers and partners are expecting us to come to them with a point of view on trends and on opportunities. And that's what differentiates us. (John laughs) That's what gives me goosebumps- >> I was just going to ask you that. Does that give you goosebumps? How could you not love technology with that excitement? I mean, AI, throw in AI, too. I just talked to Swami, who heads up the AI and database, and we just talked about the past 24 months, the change. And that is a century moment happening. The large language models, computer vision, more compute. Compute's booming than ever before. Who thought that was going to happen, is still happening? Massive change. So, I mean, if you're in tech, how can you not love tech? >> I know, even if you're not in tech, I think you've got to start to love tech because it gives you access to things you've never had before. And frankly, right, change is the only constant. And if you don't like change, you're going to like being irrelevant even less than you like change. So we've got to be nimble, we've got to adapt. And here's the great thing, once we figure it out, it changes all over again. And it's not something that's easy for any of us to operate. It's hard, right? It's hard learning new technology, it's hard figuring out what do I do next? But here's the secret. I think it's hard because we're doing it right. It's not hard because we're doing it wrong. It's just hard to be human and it's hard to figure out how we apply all this different technology in a way that positively impacts us, you know, economically, financially, environmentally and socially. >> And everyone's different, too. So you got to live those (mumbles). I want to get one more question in before we, my last question, which is about you and your impact. When you talk to your team, your sales, you got a large sales team, North America. And Tanuja, who you mentioned, is in EMEA, we're going to speak with her as well. You guys lead the front lines, helping customers, but also delivering the revenue to the company, which has been fantastic, by the way. So what's your message to the troops and the team out there? When you say, "Take that hill," like what is the motivational pitch, in a few sentences? What's the main North Star message in today's marketplace when you're doing that big team meeting? >> I don't know if it's just limited to a team meeting. I think this is a universal message, and the universal message for me is find your edge, whatever that may be. Whether it is the edge of what you know about artificial intelligence and neural networks or it's the edge of how do we migrate our applications to the cloud more quickly. Or it's the edge of, oh, my gosh, how do I be a better parent and still be great at work, right? Find your edge, and then sharpen it. Go to the brink of what you think is possible, and then force yourself to jump. Get involved. The world is run by the people that show up, professionally and personally. (John laughs) So show up and get started. >> Yeah as Steve Jobs once said, "The future "that everyone looks at was created "by people no smarter than you." And I love that quote. That's really there. Final question for you. I know we're tight on time, but I want to get this in. When you think about your impact on your company, AWS, and the industry, what's something you want people to remember? >> Oh, geez. I think what I want people to remember the most is it's not about what you've said, and this is a Maya Angelou quote. "It's not about what you've said to people "or what you've done, "it's about how you've made them feel." And we can all think back on leaders or we can all think back on personal moments in our lives where we felt like we belonged, where we felt like we did something amazing, where we felt loved. And those are the moments that sit with us for the rest of our lives. I want people to remember how they felt when they were part of something bigger. I want people to belong. It shouldn't be uncommon to talk about feelings at work. So I want people to feel. >> Rachel, thank you for your time. I know you're really busy and we stretched you a little bit there. Thank you so much for contributing to this wonderful day of great leaders sharing their stories. And you're an inspiration. Thanks for everything you do. We appreciate you. >> Thank you. And let's go do some more Women of the Cloud videos. >> We (laughs) got more coming. Bring those stories on. Back up the story truck. We're ready to go. Thanks so much. >> That's good. >> Thank you. >> Okay, this is theCUBE's coverage of International Women's Day. It's not just going to be March 8th. That's the big celebration day. It's going to be every quarter, more stories coming. Stay tuned at siliconangle.com and thecube.net here, with bringing all the stories. I'm John Furrier, your host. Thanks for watching. (gentle music)