>>LA from Las Vegas. It's the cube covering AWS reinvent 2019, brought to you by Amazon web services and along with its ecosystem partners. >>Oh, welcome back. Inside the sands, we continue our coverage here. Live coverage on the cube of AWS. Reinvent 2019. We're in day three at has been wall to wall, a lot of fun here. Tuesday, Wednesday now Thursday. Dave Volante. I'm John Walls and we're joined by Christian Rahman who was the founder and CEO of for Christian. Good morning to you. Good morning. Thanks for having afternoon. If you're watching on the, uh, on the East coast right now. Um, let's talk about sleep a little bit. I know you're all about data, um, but let's go ahead and introduce the company to those at home who might not be familiar with what your, your poor focus was. The primary focus. Absolutely. So athlete is a managed ETL as a service company. ETL is extract, transform, and load basically about getting data from different data sources, like different applications and databases into a place where it can be analyzed. >>Typically a data warehouse or a data Lake. So let's talk about the big picture then. I mean, because this has been all about data, right? I mean, accessing data, coming from the edge, coming from multiple sources, IOT, all of this, right? You had this proliferation of data and applications that come with that. Um, what are you seeing that big picture wise in terms of what people are doing with their data, how they're trying to access their data, how to turn to drive more value from it and how you serve all those masters, if you will. So there are a few trends that we see these days. One is a, you know, an obvious one that data warehouses are moving to the cloud, right? So, you know, uh, companies used to have, uh, data warehouses on premises and now they're in the cloud. They're, uh, cheaper and um, um, and more scalable, right? With services like a Redshift and snowflake in particular on AWS. Um, and then, uh, another trend is that companies have a lot more applications than they used to. You know, in the, um, in the old days you would have maybe a few data ware, sorry, databases, uh, on premises that you would integrate into your data warehouses. Nowadays you have companies have hundreds or even thousands of applications, um, that effectively become data silos, right? Where, um, uh, analysts are seeing value in that data and they want to want to have access to it. >>So, I mean, ETL is obviously not going away. I mean, it's been here forever and it'll, it'll be here forever. The challenge with ETL has always been it's cumbersome and it's expensive. It's, and now we have this new cloud era. Um, how are you guys changing ETL? >>Yeah. ETL is something that everybody would like to see go away. Everybody would just like, not to do it, but I just want to get access to their data and it should be very unfortunate for you. Right. Well, so we started, uh, we started athlete because we saw that ETL is not going away. In fact, with all the, uh, all these applications and all these needs that analysts have, it's actually becoming a bigger problem than it used to be. Um, and so, uh, what we wanted to do is basically take, take some of that pain out, right? So that companies can get to analyzing their data faster and with less engineering effort. >>Yeah. I mean, you hear this, you know, the typical story is that data scientists spend 80% of their time wrangling data and it's, and it's true in any situation. So, um, are you trying to simplify, uh, or Cloudify ETL? And if so, how are you doing that? >>So with, uh, with the growth in the number of data analysts and the number of data analytics projects that companies wants to take on the, the traditional model of having a few engineers that know how to basically make the data available for analysts, that that model is essentially now broken. And so, uh, just like you want to democratize, uh, BI and democratize analytics, you essentially have to democratize ETL as well, right? Basically that process of making the data ready for analysis. And, uh, and that is really what we're doing at athlete. We're, we're opening up ETL to a much broader audience. >>So I'm interested in how I, so I'm in pain. It's expensive. It's time consuming. Help me Christian, how, how can you help me, sir? >>So, so first of all, we're, we're, um, uh, at least specifically we're a hundred percent AWS, so we're deeply focused on, uh, Redshift data warehouses and S3 and good data lakes. Uh, and you know, there's tremendous amount of innovation. Um, those two sort of sets of technologies now, um, Redshift made a bunch of very cool announcements era at AWS reinvent this year. Um, and so what we do is we take the, uh, the infrastructure piece out, you know, so you can deploy athlete as a hosted service, uh, where we manage all the infrastructure for you or you can deploy it within your VPC. Um, again, you know, in a much, much simplified way, uh, compared to a traditional ETL technologies. Um, and then, you know, beyond that taking, uh, building pipelines, you know, building data pipelines used to be something that would take engineers six months to 18 months, something like that. But, um, but now what we, what we see is companies using athlete, they're able to do it much faster often, um, often an hours or days. >>A couple of questions there. So it's exclusively red shift, is that right? Or other analytic databases and make is >>a hundred percent AWS we're deeply focused on, on integrating well with, with AWS technologies and services. So, um, so on the data warehousing side, we support Redshift and snowflake. >>Okay, great. So I was going to ask you if snowflake was part of that. So, well you saw red shift kind of, I sort of tongue in cheek joke. They took a page out of snowflake separating compute and storage that's going to make customers very happen so they get happy. So they can scale that independently. But there's a big trend going on. I wonder if you can address it in your, you were pointing out before that there's more data sources now because of the cloud. We were just having that conversation and you're seeing the data exchange, more data sources, things like Redshift and snowflake, uh, machine intelligence, other tools like Databricks coming in at the Sage maker, a Sage maker studios, making it simpler. So it's just going to keep going faster and faster and faster, which creates opportunities for you guys. So are you seeing that trend? It's almost like a new wave of compute and workload coming into the cloud? >>Yeah, it's, it's super interesting. Companies can now access, um, a lot more data, more varied data, bigger volumes of data that they could before and um, and they want faster access to it, both in terms of the time that it takes to, you know, to, to bite zero, right? Like the time, the time that it takes to get to the first, uh, first analysis. Um, and also, um, and also in terms of the, the, the data flow itself, right? They, they not want, um, up to the second or up to the millisecond, um, uh, essentially fresh data, uh, in their dashboards and for interactive analysis. And what about the analytics side of this then when we were talking about, you know, warehousing but, but also having access to it and doing something with it. Um, what's that evolution looking like now in this new world? So lots of, um, lots of new interesting technologies there to, um, um, you know, on the, on the BI side and, um, and our focus is on, on integrating really well with the warehouses and lakes so that those, those BI tools can plug in and, and, um, um, and, and, you know, um, get access to the data straight away. Okay. >>So architecturally, why are you, uh, how are you solving the problem? Why are you able to simplify? I'm presuming it's all built in the cloud. That's been, that's kind of an obvious one. Uh, but I wonder if you could talk about that a little bit because oftentimes when we talk to companies that have started born in the cloud, John furrier has been using this notion of, you know, cloud native. Well, the meme that we've started is you take out the T it cloud native and it's cloud naive. So you're cloud native. Now what happens oftentimes with cloud native guys is much simpler, faster, lower cost, agile, you know, cloud mentality. But maybe some, sometimes it's not as functional as a company that's been around for 40 years. So you have to build that up. What's the state of ETL, you know, in your situation. Can you maybe describe that a little bit? How is it that the architecture is different and how address functionality? >>Yeah, I mean, um, so a couple of things there. Uh, um, you, you mentioned Redshift earlier and how they now announce the separation of storage and compute. I think the same is true for e-tail, right? We can, we can build on, um, on these great services that AWS develops like S three and, and, uh, a database migration service and easy to, um, elastic MapReduce, right? We can, we can take advantage of all these, all these cloud primitives and um, um, and, and so the, the infrastructure becomes operationally, uh, easier that way. Um, and, and less expensive and all, all those good things. >>You know, I wonder, Christian, if I can ask you something, given you where you live in a complicated world, I mean, data's complicated and it's getting more complicated. We heard Andy Jassy on Tuesday really give a message to the, to the enterprise. It wasn't really so much about the startups as it previously been at, at AWS reinvent. I mean, certainly talking to developers, but he, he was messaging CEOs. He had two or three CEOs on stage. But what we're describing here with, with red shift, and I threw in Databricks age maker, uh, elastic MapReduce, uh, your tooling. Uh, we just had a company on that. Does governance and, and builders have to kind of cobble these things together? Do you see an opportunity to actually create solutions for the enterprise or is that antithetical to the AWS cloud model? What, what are your thoughts? >>Oh, absolutely know them. Um, uh, these cloud services are, are fantastic primitives, but um, but enterprises clearly have a lot of, and we, we're seeing a lot of that, right? We started out in venture Bactec and, and, and got, um, a lot of, a lot of venture backed tech companies up and running quickly. But now that we're sort of moving up market and, and uh, and into the enterprise, we're seeing that they have a requirements that go way beyond, uh, beyond what, what venture tech, uh, needs. Right. And in terms of security, governance, you know, in, in ETL specifically, right? That that manifests itself in terms of, uh, not allowing data to flow out of, of the, the company's virtual private cloud for example. That's something that's very important in enterprise, a much less important than in, uh, in, in venture-backed tech. Um, data lineage. Right? That's another one. Understanding how data, uh, makes it from, you know, all those sources into the warehouse. What happens along the way. Right. And, and regulated industries in particular, that's very important. >>Yeah. I mean, I, you know, AWS is mindset is we got engineers, we're going to throw engineers at the problem and solve it. Many enterprises look at it differently. We'll pay money to save time, you know, cause we don't have the time. We don't have the resource, I feel like I, I'd like to see sort of a increasing solutions focus. Maybe it's the big SIS that provide that. Now are you guys in the marketplace today? We are. Yup. That's awesome. So how's that? How's that going? >>Yeah. Um, you mean AWS market? Yes. Yes. Uh, yeah, it's, it's um, um, that's definitely one, one channel that, uh, where there's a lot of, a lot of promise I think both. Um, for, for for enterprise companies. Yeah. >>Cause I mean, you've got to work it obviously it doesn't, just the money just doesn't start rolling in you gotta you gotta market yourselves. >>But that's definitely simplifies that, um, that model. Right? So delivering, delivering solutions to the enterprise for sure. So what's down the road for you then, uh, from, from ETL leaps perspectives here or at leaps perspectives. Um, you've talked about the complexities and what's occurred and you're not going away. ETL is here to say problems are getting bigger. What do you see the next year, 12, 18, 24 months as far as where you want to focus on? What do you think your customers are going to need you to focus on? So the big challenge, right is that, um, um, bigger and bigger companies now are realizing that there is a ton of value in their data, in all these applications, right? But in order to, in order to get value out of it, um, you have to put, uh, engineering effort today into building and maintaining these data pipelines. >>And so, uh, so yeah, so our focus is on reducing that, reducing those engineering requirements. Um, right. So that both in terms of infrastructure, pipeline, operation, pipeline setup, uh, and, and those kinds of things. So where, uh, we believe that a lot of that that's traditionally been done with specialized engineering can be done with great software. So that's, that's what we're focused on building. I love the, you know, the company tagged the perfect data pipeline. I think of like the perfect summer, the guy catching a big wave out in Maui or someplace. Good luck on catching that perfect data pipeline you guys are doing. You're solving a real problem regulations. Yeah. Good to meet you. That cause more. We are alive at AWS reinvent 2019 and you are watching the cube.

(upbeat music) >> Hey, welcome back everybody Jeff Frick here with theCUBE. We're in our Palo Alto studios, the conference season is getting ready to ramp up, it hasn't really hit full speed yet, so, it gives us the opportunity to have CUBE Conversations, and we're really excited to have our next guest, we haven't had him on for quite a while, George Mathew. He's the chairman and CEO of Kespry. George great to see you. >> Jeff, great to be here. Thanks for having me. >> So, you used to be big time in the data analytics world we used to see you at all the big data shows, and now you've made the move to autonomous flying machines. >> I did, I did, and there's a very strong relationship between the two, right? When you look at the lot worth that I was doing in the horizontal data analytic space, there was really a need to be able to accumulate data and process and understand that, and make better decisions off of it. Well, when you look at the industrial world that Kespry serves today, the ability to drive a full, complete application, where sensor based data is now being processed in our cloud infrastructure, and packaged up as complete applications, is exactly the market that we're focused on. >> So, George also a lot of big words. Let's talk about the fun words. >> Sure. >> You have drones, you have cool industrial drones. >> That's right. >> So, but what you've done is different than some the more popular drones that people know, some of the big names. You guy are really kind of single purpose, industrial only, totally integrated solution, sold as a service. >> Is that accurate? >> That's right. When you look at the drone space today, it's a big market. Its actually a 100 billion dollar market overall for drones. just in the commercial aspect of the drone space, it's a 15, 16 billion dollar market. Industrial use cases are proliferating everywhere. Kespry actually started in the mining aggregate space, where we were able to take our industrial grade drone, be able to do volumetric stock pile measurement to a level of accuracy that was literally down to one, two percent forecast accuracy, because we can now take imagery and convert that to super accurate three dimensional models of a mine site, of a query, and be able to make better decisions on how much inventory you had on that work site. >> Now, let's dive into that a little bit, cus most people when they think of drones, they think of, aerial photography at their wedding, and sweeping shots at the beach of their Maui vacation. But the industrial applications are real, and these are huge pieces of real estate that you're operating over. Huge masses of material, and men, and machines. So, the impacts, of small incremental impacts in being able to measure, and make decisions on that, have huge financial impact. >> So, what's amazing with drone tech that's available today, think about it as the new sensor network Jeff, so it's not just the fact that we can take images off a drone. It's the fact that we can take those images, and combine that with additional sensor based input. One of the key elements that Kespry introduced into the market, is taking imagery, and being able to augment the ability to have precision GPS along with that images. So, you can now have images that are processed in our cloud that are converted into full three dimensional models, and each one of those models are hyper accurate within three centimeters of real space. So, when you want to apply that for a full topological assessment of what a construction site looks like. If you wanted to measure the amount of volumetric stock pile of material that might be on 250 acres, you can fly a drone overhead in 30 minutes, be able to collect all that sensor based input, and process that in the cloud and have very accurate answers in terms of what's happening on an industrial work site without the danger and the challenges of manually collecting that information. >> Cus how did they do it before? >> Yup >> What was state of the art three years ago? >> The status quo in the market was being able to collect that data using a GPS backpack or laser guided precision equipment, but you still needed to have someone manually be able to bring that equipment to the work site. Often times, the data that you were collecting, you know, on a volumetric measurement of a stock pile, might be 20, 30, 40 points of measurement. When you're flying a drone overhead, and converting the imagery into a point cloud, you're creating five, six hundred thousand points of measurement. >> Right. >> And so the accuracy of what you're able to now accomplish with a level of safety, is unprecedented. >> Well, it's interesting, one of the Kespry tag lines is no joysticks, which I think is kind of funny. >> That's right. >> But the fact that it's really an automated system. You're selling us solutions. I'm teasing you about having fun with drones and flying with vacation, but that's not what it is. Basically it's a platform in which to deploy sensors. Which could be visual sensors, could be infrared sensors, could be GPS, could be all kinds of stuff, so it really opens up a huge opportunity to put different types of payloads, for different use cases into use. >> That's right, when you think about where Kespry's differentiation in the market is. We've introduced that capability to have different payloads, and be able to fuse those sensors together in a meaningful way, and combine that with a fully autonomous solution for flight control. So, now you don't have to have specialist piloting skills to be able to collect that information. The sensor based input is fused in a way where we can process that in our cloud infrastructure. We add a series of artificial intelligence machine learning algorithms to augment what's coming off of these sensors, and then package them as industrial grade applications. Good examples: inventory management in the mining aggregate space. Being able to do full earth works topological assessment in construction projects. Being able to do claims management for what the dimensionality, and their current state of a roof might be after a weather event has occurred. To be able to understand the number of missing shingles. The amount of hail damage that's occurred, and so all of these applications are packaged in an end to end manner, so that, you as a decision maker, and you as a user, don't have to be, you know, basically, playing with broken toys, to be able to get very clean answers in terms of what's happening in physical space. >> The roof story is so fascinating to me, 'cause people just think "oh it's a roof," they have no idea to really think through the impact of roofing in commercial real estate, and in industrial real estate. You know, roofs are where buildings fail, and so roofs, roof inspections is a really really important piece of title processes, and operational processes, so to be able now to automate that. It's classic right, automated, data driven, software driven, processes, really is a game changer versus having to send somebody up on a roof to physically inspect, I mean the accuracy's got to just be ore's of magnitude better. >> So, a few facts there, right. First of all, it's a multi billion dollar industry. You won't believe that just hail alone as far as damage that occurs on an annualized basis, is a 2.4 billion dollar challenge. It's also, the third most-- >> Is that in the U.S only? >> In the US, it's the third most occupationally hazardous job in the country, where people fall off roofs all the time when they're doing this kind of inspection. So, when you're able to now apply a drone to fly over that roof autonomously, collect that data, do the dimensional analysis, as well as being able to create the hail damage model, or the missing shingle model. You're now effectively enabling that claim process, for instance for the insurance carrier to adjudicate a claim to effectively happen within hours, right, after you know, you're on site. What we're seeing today in the market, is, if you're effectively looking at a claims assessment process, a claims adjuster would usually take about a day to cover three homes. With the use of a Kespry drone, we're seeing that same claims adjuster cover three homes in an hour. It's a massive productivity gain for this industrial use case. >> So, that brings up another topic. We've gone to a couple commercial drone shows and obviously it's a cool space, it's a fun space, but it's also really important space. I just think back to the end of World War I, when suddenly there were these things called airplanes, and the military trying to figure out, what do we do with this new asset, and those people maybe don't know that the Air Force was actually, the Army Air Force at the beginning. They didn't think that they needed a different branch, with different tactics, strategy, training, governance, et cetera. So, as we look at kind of, commercial drones entering into the business space, and I'm sure you've seen it, in some of these aggregate examples, construction. How having an air force, as a company, as a resource, you know, air deployed assets is such a big game changer. It's going to people a long time to figure out how to use it beyond the obvious in the short term, but it's a completely different tool, to apply to your business problems. >> This is why we consider this a whole new category of aerial intelligence, right. When you think about the capabilities that we're going to be able to deliver, as far as very accurate views of physical space, and being able to digitize it, to be able to model it, to be able to predict the material assets that are on a work site, and understand what the future value is, what the challenges might be for a maintenance cycle, to be able to understand the level and extent of damage, the anomaly detection, these are all incredible use cases that are opening up as we speak. I remember when I was on the show years ago, and we talked about the data analytics space, and particularly the self service aspect that I was pretty involved in, we used to talk about it being in the early innings of a ball game. Well, in the aerial intelligence market, we were literally in the first inning of the ball game. Like it is just getting off the ground, and when you think about the regulatory frameworks that are effectively in place, even as of 2016. The commercial operations in the United States have just opened up. You're now able to legitimately fly below 400 feet of air space. Maintaining the drone with a visual line of sight where a human operator is involved, that has actually passed the part 107 pilots exam. So, it's a framework. It's a start, but there's so much more expansion opportunities that occur when we're flying over people, when we're de-conflicting the air space, when we have the ability to do night flights, when we have the ability to be able to literally have that drone fly, without having a human operator controlling it, and understanding the visual line of sight where the drone is operating. So, these are all going to happen in the next several years, and completely open up the aerial intelligence market accordingly. >> It's fascinating, and of course the other thing that you're doing, which all good companies do, and all good entrepreneurs do, is build on the shoulders of others. So you're leveraging cloud, you're leveraging A.I., you're using the flight controls, you're using mobile applications, you're using all these bits and pieces of infrastructure, and you've packaged it up to deliver it as a service, which is fantastic. >> This is one of the fundamentals tenants for Kespry, even as of our founding in 2013. We knew that there was a lot of broken toys in the market, because if you had to take a consumer grade solution, be able to roll your own software, to be able to look at the way you collect that data on a manual basis, to be able to process that information, and get to results without having this connectivity involved with the entire end to end experience, we knew that a lot of companies could not succeed in their aerial intelligence offerings. And this is why Kespry believed that a full end to end solution, the way we built it, was better for the industrial markets that we serve, and so far so good. This past week we actually announced, just in the mining aggregate space alone, we have over 170 customers, and-- >> 170? >> Correct. Just in mining aggregate. >> How long has Kespry been around? >> We've been in business since we were founded in 2013. We started commercial operations in 2015. >> Wow. >> Amazingly, we covered over 10,400 just, mining query work sites, just in those last two and a half years that we've been in commercial operation. So, this is something that has really exponentialized, just in that market, and we're seeing similar adoptions starting to take off in the insurance roofing space, as well the construction markets. >> It's so funny. I just consider, you're an autonomous vehicle. You're just one that flies, not, that drives on the road, but, there's so much going on on the commercial side that people don't see, you know? They see the Lambo cars driving around the neighborhood, and we read about what's going on with Tesla, but on the agg side, on the commercial side, with John Deere, and these huge mining trucks, that many of them are already autonomous. This stuff is really moving very very quickly on the commercial side. >> If you think about the digital transformation of industrial work. This is a one trillion dollar market opportunity over the next several decades, and the ability to sense physical assets, and be able to make better decisions using drone tech, using other sensor based information. This is transforming the nature of industrial work, right? This is, in my view, the beginning of the fourth industrial age, and in that regard, we see this as something that's not just, like I said, you know, the first few innings of a ball game. We're going to see this evolve for decades, as we move forward. And drones are effectively a critical piece of that infrastructure evolving. >> Yeah, just in delivery. Just sensor delivery is basically what it is. Place it in places that people maybe shouldn't go, don't want to go, that dangerous to go, it makes a ton of sense. >> And then being able to blend that with the other sensors that might be on the ground, that might be in other places, that you can fuse that information together to get better understanding of physical space. >> Yeah, I love it. I love the solution approach, right. Nobody ever buys a new platform, but it sure is great to build a platform underneath a terrific application, that then you can expand after you knock it out of the park with that first application. >> And that's exactly the approach that we're going after >> All right. Well Mat, hopefully it won't be a, we looked it up before. Last time you were on was like 2014, so hopefully-- >> It's been a while. >> It won't be so long before we see you next, and thanks for stopping by. >> Thanks for having me on board, Jeff. >> All right, he's George Mathew. I'm Jeff Frick, You're watching theCUBE. Thanks for watching, I'll see you next time. (upbeat music)

>> Voiceover: Live from New Orleans, it's the Cube covering VeeamON 2017 brought to you by Veeam. >> Welcome back to New Orleans everybody. This is the Cube, the leader in live tech coverage. We go out to the events and extract the signal from the noise. My name is Dave Vellante, and I'm here with my cohost Stu Miniman. Frank Palumbo is here. He's the senior vice president at Cisco Systems. And Andy Vandeveld is the vice president of Global Alliances at Veeam Software. Gents, welcome to The Cube. >> How we doing? >> Thank you. >> It's great to be here. >> Good, Frank, hot off the keynote. It was great, Yankees fan, love it. The rivalry continues. Of course you guys know the Cube, Red Sox fans, some of us. Stu's not. >> Not all of us. >> So we love it. We love the action, and it's always fun. But Frank we had to cut out a little bit before your keynote because we had to get ready to do the Cube. But you put up a slide that was awesome. We could do an hour on The Cube on that, and it's all about the apps, I mean really. But you had this great slide with apps and microservices and virtualization and bare metal and OnPrim and really laying out the complexity today. And you guys are at the heart of that. Maybe give us a quick summary of how you guys see the world. >> When you're talking about the applications, the application profile, it's important, the network kind of brings this together because we do touch everything. Where people are in this kind of application history is some of them are on legacy, mainframe. Some of them are on RISC processors. But as a network provider, we have to bring those in too even with the more modern applications. So you look at what the platforms or workloads are on so move those in. And then you're looking at workload placement, on Prim or in the Cloud. Do we put data in a colo? Do we put the application in the Cloud? There's different hybrid mentalities to do that. Then you get into the systems management where there's just too much stuff out there. Humans can't manage it anymore so the machines and the software have to manage the machines and the software. We'd like to think we're right in the middle of that because of the way we bring things together with the network. >> So Andy, I look at the... Stu and I walked the floor before we come in here, the ecosystem is really quite impressive-- >> Andy: Thank you. >> for a relatively small company. I mean not that small anymore. It didn't just happen overnight. Maybe you could talk a little bit about themes and philosophy with partnerships and some of the things that you're doing with Alliances generally and specifically get into the Cisco partnership. >> Well I think partnerships have been in our DNA since the beginning of the company. We're a 100% channel-lead company. We don't have a direct sales force. That's an important piece of the company's philosophy. These alliances are really key for us because as we start to move into markets that are maybe a little bit higher than where we've been into the large enterprise and mid-enterprise and large enterprise, we really look at partnerships like the one with Cisco that are going to benefit Veeam and the customers by us being together doing joint developments. Some of the things that Frank talked about in his keynote speech, those are the sorts of things that create solutions for that level of customer where Cisco's been resident for many, many years. So we look at these partnerships as really central to where Veeam wants to go as a company and where we think customers want Veeam to participate with the partners. >> What's the specific nature of the partnership? Can you unpack that a little bit for us? >> From my side, certainly we have a robust go-to-market relationship in terms of when we're positioning UCS or Hyperflex, our server and hyper converged platforms, now we can bring to bear the Veeam value problem as we go forward with customers. And customers look to Cisco really to complete the story and offer an end-to-end solution. We weren't able to complete it without the Veeam technology. Then on the development side, some of the things that we're doing, we've integrated so now the Veeam software can work with our Snap technology and hyper converge. So you're starting to see it come together at the screen level with the bits and bytes in terms of the integration. >> Dave: So there's a greater degree of technical integration as well. >> Frank: Yes. >> It's not just go-to, I mean that's important because a lot of times back-up data protection is kind of an afterthought. It's a bolt-on. But if you're going to be a complete solution provider, that's fundamental and it's becoming more important. >> I think you know I was just mentioning to Frank back in the green room before we came out here I look at the start of this partnership as really being about 18 months ago. Although we'd had a partnership for a while, we really kind of started about 18 months ago in this meeting that we had at their partner conference in Maui. And Radmeer and I sat down with Frank and kind of explained why we thought data protection was a solution that Cisco could get behind particularly now that they were coming out with their S-Series devices. But that's just the start of it. It has to come with integration as well. Then we started with Hyperflex. It was a new product for them, 1.0 version. With the 2.0 version, we got integrated with snapshot technology like Frank mentioned. I look at this short runway of time in this relationship that kicked off with our meeting with Frank and he got it right away. We didn't have to explain it. >> Dave: It resignated. >> Frank: Oh, no question. We're very proud of our S-Series storage server. The hardware is nice. The infrastructure piece is nice, but it really doesn't come together unless you got the application on a run with it. That's where Veeam just jumps in and fills that gap perfectly for us. >> Frank, I think back to when virtualization really took off. Networking was one of the things that we had to fix. It put a lot of stress on the network. It's one of the reasons Cisco created UCS and backup also creates a lot of strain on the network. So it seems a natural fit. Can you talk about all the complexities that are coming and how you're going to be, what can we expect to see from jointly going forward? >> I think we've learned a lot from Veeam in terms of they've been able to really attack complex issues in a very simple fashion. Simplicity is the game with customers right now. Things are moving so fast. If you can't be simple, you're going to have a tough time out there. So I think that's where it's really come together for us in that vein. But when you look at the value of data and whether it's a second old or two years old or a year old, there's so many different more paradigms coming out about what you can do with this data. And customers and even customers of customers have now found ways to use this data either to make better decisions, monetize it, to stay away from things. So that's why this whole lifecycle for us is so important. This is where Veeam and us can really do some nice things for customers. >> Andy, can you build on that about the multi-Cloud position that Veeam has? How many of those, do you know, touch what Cisco's doing here and how does the partnership help drive that value of data type offering? >> For Veeam, our message is all about availability, availability of the data which makes the applications available and which basically makes the business stay up and running. One analogy we use is a cell phone. When you're cell phone dies, you can't get access to your email. You can't get access to your instant messages. >> Dave: You freak bascially. >> You feel like you're lost, right? >> Frank: It's getting kind of pathetic. >> Yeah. >> Dave: It is pretty bad. >> So think about not being able to get access to your data or access to your applications because of some outage, not being able to backup and recover. Your business could go out of business. Working with Cisco on solutions that are on premise, that are in the Cloud, that are multi-Cloud is really the value of the partnership that we have that we bring together. It's just at the beginning. We've got solutions that we're building now. We got solutions that are on the horizon. We've got a very strong go-to-market partnership in a very short period of time that are targeting enterprise customers, service providers, the whole gamut. It's really that sort of relationship that you find in an industry every so often. When it comes together like it has with us and Cisco, it's really a very strong, strong value prop. >> Well Veeam capitalized on the original virtualization trend with VMware that was a big transformation, the server infrastructure. You're seeing a huge network transformation now. There are so many forces affecting the network that I wonder, Frank, if you could comment on. You got ScaleOut. There's Flash. There's Cloud. There's Microservice. There's DevOps makes everything go faster. The flattening of the network. Describe what's happening and then maybe you can talk about how your ecosystem is going to take advantage of that. From what were the challenges the network has is exactly like you said. You have certainly the virtualized workloads now. The Microservices containerize workloads. I think the one people forget about is there's still a ton of bare metal out there, right? You look at the Hadoop workloads and such. A lot of these are bare metal oriented, right? Quite frankly, moving a VM around a fabric is actually pretty easy to do. But when you got to move a bare metal workload around a fabric, and that's something we can do with UCS the way we do it statelessly, that's much harder. That's why we have the extraction layer with what we call the fabric interconnection with UCS to do that kind of stuff. I think that's sometimes lost in the translation in terms of how you're going to handle all these different workloads. >> If I understand it, the link then to the opportunity for you guys, Andy, is that the stakes are just much higher now, right? You could do so much more around the networks. Stakes are so much higher. That increases the need for your products and services. Carry that through if you would. >> Well, it is. As we make our way up-market into the enterprise, the amount of data that businesses are spinning off of, their infrastructure and their data center or from robo offices or wherever, is growing immensely. Being able to have a partnership with an infrastructure provider like Cisco, where we can put solutions together that really give the customers the rock solid base for backing up their data and making sure that it's available is really critical for us as we move into those larger enterprise and larger environments. So this is an essential relationship I would say. >> I think, too, if I could mention, this is something our channel wanted to see, too. We're the same. We're at about 98% of our business goes through the channel. So they're selling our full line of infrastructure products. This completes the story for them. So we got a lot of guides to them say, "Hey, yes, Cisco. "We'd like to see you come together with Veeam "so we can start bundling offers out there in the market "and be that kind of end-end-to supplier, too." That was a big impetus especially from mid-market up to enterprise customers. >> Excellent, well, we got to wrap there. The partnerships give you huge leverage as a small, again not so small company anymore. The fact that you can get somebody like Frank to come down, talk about the partnership, is a testament to what you guys have built. So congratulations. Really appreciate you guys coming on The Cube. >> No, my pleasure, our pleasure. >> All right, keep it right there, everybody. We'll be back with our next guest. This is The Cube. We're live from New Orleans, VeeamON 2017. We'll be right back. (tinkling music)

why from the mandalay bay convention center in las vegas it's the cues covering vmworld 2016 rock you buy vmware and its ecosystem sponsors now you're your host John furrier and John wall's well welcome back to Mandalay Bay here at vmworld along with John furrier I'm John wall's glad to be with you here on the cubes to continue our coverage what's happening at vmworld exclusive broadcast a partner here for the show and along with John we're joined by eric Herzog's the vice president product marketing and management at IBM storage and Erica I just found out you're one of the all-time 10 most popular cute guests or most prominent cube guests most prolific congratulations well thank you we always love coming to the cube it's always energizing you love controversy and I love controversy and you get down to the heart of it you're the hard copy of high tech they're like oh I loved and we could probably mark each of your appearances by the Hawaiian shirt I think what do you think either Hawaiian shirt or one of my luggage share we could trace those back ever stop vibe about the show I mean just your thoughts about they've been here for three four days now just your general feel about about the the messaging here and then what's actually being conveyed in the enthusiasm out on the show floor well it's pretty clear that the world has gone cloud the world is doing cognitive and big data analytics vmware is leading that charge their strong partner of IBM we do a lot of things with them both with our cloud division on our storage division and vmware is a very strong partner of IBM we have all kinds of integration in our storage technology products with vai with vasa with vcenter ops all the various product lines at vmware offers and the key thing is ever wants to go to the cloud so by working with IBM and vmware together makes it easier and easier for customers whether it be the small shop Herzog's barn grill or whether it be the giant fortune 500 global entity working with us together allow them to get to the cloud sooner faster and have a better cloud experience so you got you know everybody cloud and virtualization and you know big themes big big topics so why does storage still matter well the big thing is if you're going to go to a cloud infrastructure and you're going to run everything on the cloud you think of storage as at solid foundation it has to be rock solid it has to be highly resilient it has to be able to handle error codes and error messaging and things failing and things falling off the earth at the same time it needs to be incredibly fast where things like all-flash arrays come in and even flexible so things like software-defined storage so think of storage as the critical foundation underneath any cloud or virtualized environment if you don't have a strong storage foundation with great resiliency great availability great serviceability and great performance your cloud or your virtual infrastructure is going to be mediocre and that's a very generous term so that's a key point so controversial II speaking to get to the controversy the whole complexity around converged infrastructure hyper converge or whatever the customers are deploying for compute they're putting the storage close to that whether it's a SAS and the cloud which is basically a data center that no one knows the address of as we were saying they always going to have stores has to sit somewhere what is the key trends right now for you because software is leading the way iBM has been doing a lot of work I know and soft we've been covering you guys will be at IBM edge coming up shortly in a couple weeks where's the innovation on the storage side for you guys well how do you talk to the customer base to say ok I got some sass options now for back and recovery weird one of your partners earlier i'm talking about that where is the physical storage innovation is that the software what's your thoughts on so we have a couple paths of integration for us first software-defined storage several the other analyst firms have named it's the number one software-defined storage coming in the world for several years in a row now software-defined storage gives a flexible infrastructure you don't have to buy any of the underlying media or underlying array controller from us just by our software and then you could put on anybody else's hardware you want you can work with your cloud provider with your reseller with your distributor enterprises create their own cloud whether it's a software-defined storage gives you a wide swath of storage functionality backup archive primary store grid scale out software only so ultimate flexibility so that one area of innovation secondary ish is all flash all flash is not expensive essentially I love old Schwarzenegger movies in the 1980s was all about tape he was a spy go and show what is supposedly the CIA was Schwarzenegger I'll take mid 90s Schwarzenegger another spy movie show a datacenter all hard drive arrays now in the next Schwarzenegger movie hopefully it'll be all flash arrays from IBM in the background so flash is just an evolution and we do tons of humor white shirts I keep swapping monitors it so he's intimated I get one from Maui went from kawaii one from the Big Island so flash is where it's at from a system level perspective so you've got that innovation and then you've got converged infrastructure as you mentioned already will you get the server the storage the networking and VMware hypervisor all packaged up dramatically so we have a product called the vs tak we do jointly with Cisco and vmware we were late to market on that we freely admit that but just give you an idea in the first half of this year we have done almost 2x what we did in the entire year of 2015 so that's another growth ending particularly cloud service providers love to get these pre-canned pre racked versus tax and deploy them in a number of our public references are cloud service providers both big and small essentially wheel in a versus stack when they need it whelan not own will another pre-configured ready to go and they get up and up and quit going so those are three trends we just had a client on Scott equipment not a Monroe Louisiana went to the Versa stack and singing your praises like a great example of medium size small sized businesses so we keep think about enterprises and all this and that it doesn't have to be the case their services that you're providing the companies of all sizes that are gaining new efficiencies in protocol al people everybody needs storage and you think about it is really how do you want to consume the storage and in a smaller shop you may choose one way so versus stack is converged infrastructure our software-defined storage like spectrum accelerate spectrum virtualize a software-only model several of the products like spectrum accelerate inspect can protect are available through softlayer or other cloud is he consumed it as a cloud entity so whether you want to consume an on-premises software only full array full integrated stack or cloud configuration we offer any way in which you want to eat that cake big cake small cake fruit cake chocolate cake vanilla cake we got kicked for ever you need and we can cover every base with that a good point about the diversity of choices from tape to flash and they get the multi multi integrated Universal stack so a lot of different choices I want to ask you about you know with that kind of array of options how you view the competitive strategy for IBM with storage so you know I know you're a wrestler so is there a is there a judo move on the competition how would you talk about your differentiation how do you choke hold the competition well couple ways first a lot from a technical perspective by leading with software-defined storage and we are unmatched in that capacity according the industry analysts on what we do and we have it in all areas in block storage we got scale-out file storage and scale out big data analytics we got back up we got archive almost no one has that panoply of offering in a software-defined space and you don't need to buy the hardware from us you can buy from our competitors two things I hear software and then after the array of eyelash what's specifically on the software are you guys leading and have unmatched as-safir already well spectrum protect is you know been a leader in the enterprise for years spectrum scale is approaching 5,000 customers now and we have customers close to an exabyte in production single customer with an exabyte pretty incredible so for big data analytic workloads with on gastronomic research so for us it's all about the application workload in use case part of the reason we have a broad offering is anyone who comes in here and sits in front of you guys and says my array or my software will do everything for you is smoking something that's not legal just not true maybe in Colorado or yeah okay me but the reality is workloads applications and use cases very dramatically and let's take an easy example we have multiple all-flash arrays why do we have multiple all flash arrays a we have a version for mainframe attached everyone in there wants six or seven 9s guess what we can provide that it's expensive as they're all is that our six or seven 9s but now they can get all flash performance on the mainframe in the upper end of the Linux world that's what you would consume at the other end we have our flash our store wise 50 30 f which can be as low street price as low street price as eighteen thousand dollars for an all-flash array to get started basically the same prices our Drive rang and it has all the enterprise data services snapshot replication data encryption at rest migration capability tiering capability it's basically what a hard drive array used to cost so why not go all flash threat talk about the evolution of IBM storage actually them in a leader in storage in the beginning but there was a period of time there and Dave when I won't talk handling the cube about this where storage my BMC it took a lot of share but there's been a huge investment in storage over the past i'd say maybe five years in particular maybe past three specifically i think over a billion dollars has been spent I think we thought the Jamie talent variety of folks on from IBM what is the update take a minute to explain how IBM has regained their mojo in storage where that come from just add some color to that because I think that's something that let people go hmm I great for things from my being but they didn't always have it in the storage so as you know IBM invented the hard drive essentially created the storage industry so saying that we lost our mojos a fair statement but boy do we have it back explain so first thing is when you have this cloud and analytic cognitive era you need a solid foundation of storage and IBM is publicly talked about the future of the world is around cloud on cognitive infrastructure cognitive applications so if your storage is not the best from an availability perspective and from a performance perspective then the reality is your cloud and cognitive that you're trying to do is basically going to suck yeah so in order to have the cloud and convey this underlying infrastructure that's rock-solid so quite honestly as you mentioned Dave we've actually invested over three and a half billion dollars in the last three years not to mention we bought a company called Texas memory systems which is the grandfather our flash systems knocks before that so we've invested well over three billion dollars we've also made a number of executive hirings ed walls just joined us CEO of several startups former general manager from emc i myself was a senior vice president at emc we just hired a new VP of Sales they're serious you guys are serious you guys are all in investing bringing on the right team focusing on applications work gloves in use case as much as I love storage most CEOs hate it yeah there's almost no cio that whatever a storage guy they're all app guys got to talk their lingo application workload in use case how the storage enables their availability of those apps workloads and use cases and how it gives them the right performance to meet their essays to the business guy what's interesting I want to highlight that because I think it's a good point people might not know is that having just good storage in and of itself was an old siloed model but now you mentioned could we cover all the IBM events world of Watson we should call insights edge and and interconnect the cloud show cognitive is front and center there's absolutely the moon shot and the mandate from IBM to be number one in cognitive computing which means big data analytics integrated to the application level obviously bluemix in the cloud Philip blank was here on stage about IBM cloud the relation with VMware so that fails if it doesn't have good steward doesn't perform well and and latency matters right I mean data matters well I add a couple things there so first of all absolutely correct but the other thing is we actually have cognitive storage ok if you automate processes automatically for example to your data some of our competitors have tiering most of them tier only within their own box we actually can tear not only within our own box for from our box to emc our box to netapp our box to HP HP to del Delta hitachi we can t r from anything to anything so that's a huge advantage right there but we tier we don't just set policy which is when data's 90 days old automatically move it that's automation cog nation is where we not only watch the applications and watch the data set we move it from hot to cold so let's take for example financial data your publicly traded company cuban SiliconANGLE going to be public soon i'm sure guys are getting so big your finance guys going to say Dave John team this financial data is white-hot got to be on all flash after you guys do your announcement of your incredible earnings and thank God I hopefully get friend of the company stock and my stock goes way up as your stock goes way up what are we spoking now come on let me tell you when that happens the date is going to go stone-cold we see that you don't have to set a policy two-tier the data with IBM we automatically learn when the data is hot and when it's cold and move it back and forth for you you know there's no policy setting cognition or cognitive its storage understand or stands out as the work for some big data mojo coming into the storage right and that's a huge change so again not only is it critical for any cognitive application to have incredibly performance storage with incredible resiliency availability reliability ok when there is cognitive health care true cognitive health care and Dave's on the table and they bring out their cognitive Juan because they found something in your chest that they didn't see before if the storage fails not going to be good for Dave yeah at the same time if the storage is too slow that might not be good for Dave either when they run that cognitive wand a that hospital knows that it's never going to fail that doctor says Oh Dave okay we better take that thing out boom he takes it out Dave's healthy again well that's a real example by the way not necessary Dave on the table but there was a story we wrote insult an angle one of our most popular post last month IBM Watson actually found a diagnosis uncured a patient the doctor had missed I don't know if you saw that story when super viral but that's the kind of business use case that you're in kind of illuminating with the storage yeah well in fact that one of the recent trade shows what's called the flash memory summit we won an award for best enterprise application commercial developer spark cognition they developed cyber security applications they recommend IBM flash systems and actually Watson's embedded in their application and it detects security threats for enterprises so there's an example of combining cognition with Watson the cognition capability of flash systems and then their software which is commercially available it's not an in-house thing or they're you know a regular software all right now we're a now we're in like the big time you know intoxication mode with all this awesome futuristic real technology how does a customer get this now because now back to IT yeah the silos are still out there they're breaking down the silos how do you take this to customers what's to use case how do you guys deploy this what's the what are you seeing for success stories well the key thing is to make it easy to use and deploy which we do so if you want the cloud model we're available in software IBM Global resiliency services uses us for their resiliency service over 300 cloud providers you spectrum protect for backup pick the cloud guy just pick one you want we work with all of them if you want to deploy in-house we have a whole set of channel partners globally we have the IBM sales team IBM global services uses IBM's own storage of course to provide to the larger enterprises so with your big shop medium swaps well flop we have a whole set of people out there with our partner base with our own sales guys that can help that and you get up and then we back it up as you know IBM is renowned for supporting service in all of our divisions in all of our product portfolio not just in storage so they need support and service our storage service guys are there right away you'd it installed we can install it our partners can install this stuff so we try to make it as brain dead as possible as easy as possible Jen being cognitive and are some of our user interfaces are as easy as a Macintosh I mean drag-and-drop move your lungs around run analytics on when you're going to run out of storage so you know ahead of time all these things that cut things people want today remember IT budget cut dramatically in the downturn of 08 09 and while budgets have returned they're not hiring storage guys there are hiring developers and they're hiring cloud guys so those guys don't know how to use storage well you got to make it easy always fast and always resilient that way it doesn't fail anyway but when it does you just go into the GUI it tells you what's wrong bingo and IBM service our partner service comes right out and fix it so that's what you need today because there aren't as many storage guys as you used to be no question you've got the waterfront covered no doubt about that and again congratulations on cracking the top 10 way back we consider that an honor and a privilege to be a part of that great welcome picture we really appreciate it thank you we'll continue the coverage here on the Cuba vmworld right after this

>>The cube at IBM. Impact 2014 is brought to you by headline sponsor. IBM. Here are your hosts, John furrier and Paul Gillin. Okay, welcome back. Everyone live in Las Vegas at IBM impact. This is the cube, our flagship program. We go out to the events, instruct us to live in the noise. I'm John Ferrari, the founder of SiliconANGLE Joe, my close Paul Gillen. And our next special guest is great bushes as a legend in the software development community. And then she went to st this school in Santa Barbara. My son goes there, he's a freshman, but there's a whole nother conversation. Um, welcome to the cube. Thank you. Uh, one of the things we really exciting about when we get all the IBM guys get the messaging out, you know, the IBM talk, but the groundbreaking work around, um, computer software where hardware is now exploding and capability, big data's instrumentation of data. >>Um, take us to a conversation around cognitive computing, the future of humanity, society, the societal changes that are happening. There's a huge, uh, intersection between computer science and social science. Something that's our tagline for Silicon angle. And so we are passionate about. So I want to, I just want to get your take on that and, and tell about some of the work you're doing at IBM. Um, what does all this, where's all this leading to? Where is this unlimited compute capacity, the mainframe in the cloud, big data instrumentation, indexing, human thought, um, fit, Fitbit's wearable computers, um, the sensors, internet of things. This all taking us in the direction. What's your vision? There are three things that I think are inevitable and they're irreversible, that have unintended consequences, consequences that, you know, we can't, we have to attend to and they will be in our face eventually. >>The first of these is the growth of computational power in ways we've only begun to see. The second is the development of systems that never forget with storage beyond even our expectations now. And the third is a pervasive connectivity such that we see the foundations for not just millions of devices, but billions upon billions of devices. Those three trends appear to be where technology is heading. And yet if you follow those trends out, one has to ask. The question is you begin to, what are the implications for us as humans? Um, I think that the net of those is an interesting question indeed to put in a personal blog. My wife and I are developing a documentary or the computer history with the computer history museum for public television on that very topic, looking at how computing intersects with the human experience. So we're seeing those changes in every aspect of it too, that I'll dwell upon here, which I think are germane to this particular conference are some of the ethical and moral implications. >>And second, what the implications are for cognitive systems. On the latter case we saw on the news, I guess it was today or yesterday, there's a foundation led by the Gates foundation. It's been looking at collecting data for kids in various schools. A number of States set up for it. But as they begin to realize what the implications of aggregating that information were for the privacy of that child, the parents became, became cognizant of the fact that, wow, we're disclosing things for which there can be identification of the kid in ways that maybe we wouldn't want to do that. So I think the explosion of big data and explosion of computational power has a lot of us as a society to begin asking those questions, what are the limits of ownership and the rights of that kind of information. And that's a dialogue that will continue on in the cognitive space. >>It kind of follows on because one of the problems of big data, and it's not just you know, big, big data, but like you see in at CERN and the like, but also these problems of aggregation of data, there are, there are such an accumulation information at such a speed in ways that an individual human cannot begin to reason about it in reasonable ways. Thus was born. What we did with Watson a few years ago, Watson jeopardy. I think the most important thing that the Watson jeopardy experience led us to realize is that theory is an architectural framework upon which we can do many interesting reasoning things. And now that Watson has moved from research into the Watson group, we're seeing that expand out in so many domains. So the journey is really just beginning as we take what we can know to do in reason with automated systems and apply it to these large data systems. >>It's going to be a conversation we're going to have for a few generations. You were beginning to see, I mean computing has moved beyond the, the, the role of automate or of automating rote manual tasks. We're seeing, uh, it's been, uh, I've seen forecast of these. Most of the jobs that will be automated out of existence in the next 20 years will be, will be, uh, knowledge jobs and uh, even one journalism professor of forecasting, the 80% of journalism jobs will go away and be replaced by computer, uh, over the next couple of decades. Is this something for people to fear? I'm not certain fear will do us any good, especially if the change like that is inevitable. Fear doesn't help. But I think that what will help is an understanding as to where those kinds of software systems will impact various jobs and how we as individuals should relate to them. >>We as a society, we as individuals in many ways are slowly surrendering ourselves to computing technology. And what describe is one particular domain for that. There's been tremendous debate in the economic and business community as to whether or not computing has impacted the jobs market. I'm not an economist, I'm a computer scientist, but I can certainly say from my input inside perspective, I see that transformational shift and I see that what we're doing is radically going to change the job market. There was, you know, if you'd go back to the Victorian age where people were, were looking for a future in which they had more leisure time because we'd have these devices to give us, you know, free us up for the mundane. We're there. And yet the reality is that we now have so many things that required our time before. It means yours in a way, not enough work to go around. >>And that's a very different shift than I think what anyone anticipated back to the beginnings of the industrial age. We're coming to grips with that. Therefore, I say this, don't fear it, but begin to understand those areas where we as humans provide unique value that the automated systems never will. And then ask ourselves the question, where can we as individuals continue to add that creativity and value because there and then we can view these machines as our companions in that journey. Great. You want to, I want to ask you about, um, the role, I mean the humans is great message. I mean that's the, they're driving the car here, but I want to talk about something around the humanization piece. You mentioned, um, there's a lot of conversations around computer science does a discipline which, um, the old generation when a hundred computer science school was, it was code architecture. >>But now computer science is literally mainstreams. There's general interest, hence why we built this cube operation to share signal from the noise around computer science. So there's also been a discussion around women in tech tolerance and different opinions and views, freedom of speech, if you will, and sensors if everything's measured, politically correctness. All of this is now kind of being fully transparent, so, so let's say the women in tech issue and also kids growing up who have an affinity towards computer science but may not know us. I want to ask you the question. With all that kind of as backdrop, computer science as a discipline, how is it going to evolve in this space? What are some of those things for the future generation? For the, my son who's in sixth grade, my son's a freshman in college and then in between. Is it just traditional sciences? >>What are some of the things that you see and it's not just so much coding and running Java or objective C? I wish you'd asked me some questions about some really deep topics. I mean, gosh, these are, these are, I'm sorry. It's about the kids. In the early days of the telephone, phone, telephones were a very special thing. Not everybody had them and it was predicted that as the telephone networks grew, we were going to need to have many, many more telephone operators. What happened is that we all became, so the very nature of telephony changed so that now I as an individual have the power to reach out and do the connection that had to be done by a human. A similar phenomenon I think is happening in computing that it is moved itself into the interstitial spaces of our world such that it's no longer a special thing out there. We used to speak of the programming priesthood in the 60s where I lost my thing here. Hang on. >>Here we go. I think we're good. We're good. I'm a software guy. I don't do hardware so my body rejects hardware. So we're moving in a place where computing very much is, is part of the interstitial spaces of our world. This has led to where I think the generation after us, cause our, our median age is, let me check. It's probably above 20, just guessing here. Uh, a seven. I think you're still seven. Uh, we're moving to a stage where the notion of computational thinking becomes an important skill that everyone must have. My wife loves to take pictures of people along the beach, beautiful sunset, whales jumping and the family's sitting there and it did it again. My body's rejecting this device. Clearly I have the wrong shape. i-Ready got it. Yeah. There we go. Uh, taking pictures of families who are seeing all these things and they're, they're very, with their heads in their iPhones and their tablets and they're so wedded to that technology. >>We often see, you know, kids going by and in strollers and they've got an iPad in front of them looking at something. So we have a generation that's growing up, uh, knowing how to swipe and knowing how to use these devices. It's part of their very world. It's, it's difficult for me to relate to that cause I didn't grow up in that kind of environment. But that's the environment after us. So the question I think you're generally asking is what does one need to know to live in that kind of world? And I think it says notions of computational thinking. It's an idea that's come out of uh, the folks at Carnegie Mellon university, which asks the question, what are some of the basic skills we need to know? Well, you need to know some things about what an algorithm is and a little bit behind, you know, behind the screen itself. >>One of the things we're trying to do with the documentary is opening the curtain behind just the windows you say and ask the question, how do these things actually work because some degree of understanding to that will be essential for anyone moving into, into, into life. Um, you talked about women in tech in particular. It is an important question and I think that, uh, I worked with many women side by side in the things that I do. And you know, frankly it saddens me to see the way our educational system in a way back to middle school produces a bias that pushes young women out of this society. So I'm not certain that it's a bias, it's built into computing, but it's a bias built in to culture. It's bias built into our educational system. And that obviously has to change because computing, you know, knows no gender or religious or sexual orientation boundaries. >>It's just part of our society. Now. I do want to, everyone needs to contribute. I'm sorry. I do want to ask you about software development since you're devoted your career to a couple of things about to defining, uh, architectures and disciplines and software development. We're seeing software development now as epitomized by Facebook, perhaps moving to much more of a fail fast mentality. Uh, try it. Put it out there. If it breaks, it's okay. No lives were lost. Uh, pull it back in and we'll try it again. Is this, is there a risk in, in this new approach to software? So many things here are first, is it a new approach? No, it's part of the agile process that we've been talking about for well over a decade, if not 15 years or so. You must remember that it's dangerous to generalize upon a particular development paradigm that's applied in one space that apply to all others. >>With Facebook in general, nobody, no one's life depends upon it. And so there are things that one can do that are simplifying assumptions. If I apply that same technique to the dialysis machine, to the avionics of a triple seven, a simple fly, you know, so one must be careful to generalize those kinds of approaches to every place. It depends upon the domain, depends upon the development culture. Ultimately depends upon the risk profile that would lead you to high ceremony or low ceremony approaches. Do you have greater confidence in the software that you see being developed for mission critical applications today than you did 10 years ago? Absolutely. In fact, I'll tell you a quick story and I to know we need to wind down. I had an elective open heart surgery or a few years ago elective because every male in my family died of an aneurysm. They are an aneurism. >>So I went in and got checked and indeed I had an aneurysm developing as well. So we had, you know, hi my heart ripped open and then dealt with before it would burst on me. I remember laying there in the, in the, uh, in the CT scan machine looking up and saying, this looks familiar. Oh my God, I know the people that wrote the software where this thing and they use the UML and I realized, Oh this is a good thing. Which is your creation. Yes. Yes. So it's a good thing because I felt confidence in the software that was there because I knew it was intentionally engineered. Great. I want to ask you some society questions around it. And computing. I see green as key and data centers take up a lot of space, right? So obviously we want to get to a smarter data center environment. >>And how do you see the role of software? I see with the cognitive all things you talked about helping businesses build a physical plant, if you will. And is it a shared plan is a Terminus, you seeing open power systems here from IBM, you hear him about the open sources source. Um, what, what does that future look like from your standpoint? May I borrow that cup of tea or coffee? I want to use it as a aid. Let's presume, Oh, it's still warm. Let's say that this is some tea and roughly the energy costs to boil water for a cup of tea is roughly equivalent to the energy costs needed to do a single Google search. Now imagine if I multiply that by a few billion times and you can begin to see the energy costs of some of the infrastructure, which for many are largely invisible. >>Some studies suggest that computing is grown to the place releasing the United States. It's consuming about 10% of our electrical energy production. So by no means is it something we can sweep under the rug. Um, you address I think a fundamental question, which is the hidden costs of computing, which believe people are becoming aware of the meaning. Ask the question also. Where can cognitive systems help us in that regard? Um, we live in, in Maui and there's an interesting phenomenon coming on where there are so many people using solar power, putting into the power grid that the electrical grid companies are losing money because we're generating so much power there. And yet you realize if you begin to instrument the way that people are actually using power down to the level of the homes themselves, then power generation companies can start making much more intelligent decisions about day to day, almost minute to minute power production. >>And that's something that black box analytics would help. But also cognitive systems, which are not really black box analytic systems, they're more learn systems, learning systems can then predict what that might mean for the energy production company. So we're seeing even in those places, the potential of using cognitive systems for, for uh, attending to energy costs in that regard. The future is a lot of possibilities. I know you've got to go, we're getting the hook here big time cause you gotta well we really appreciate it. These are important future decisions that are, we're on track to, to help solve and I really appreciate it. Looking for the documentary anytime table on that, uh, sometime before I die. Great. Thanks for coming on the, we really appreciate this. This SiliconANGLE's we'll be right back with our next guest at to nature. I break.