>>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. >>Welcome to the cube at Lisa Martin with Dave Volante. This is our first day of covering AWS reinvent 2019 Dave, we have a jam packed three days here. The seventh time the cube has been at reinvent the super Superbowl. Here it is. I, I co I stole that from you but you just send it back to me. It is like the super bowl here. We're very pleased to welcome a couple of guests from refitted refinished tips, first time on the cube as well as our guest. Please welcome Hannah. We've got Hannah Helen, Helen's, our VP of cloud propositions and James Turk, the head of architecture and cloud from refinish. Guys, welcome to the cube. You. Thank you for having us. So here we are in the expo hall with thousands and thousands of folks, but I'd love for you guys to start a Hannibal. Start with you. Tell our audience about refinish if you're a data company, but really what is it that you guys do? What do you deliver to the community? Absolutely >>what we are, as I said, we are a data company, so we serve the global financial community. So we're looking at banks, asset managers, hedge funds, corporations with financial and risk data. That's a very powerful combination in these clouds. Environmental or we say without data flower is empty. So that's where we come in. >>And what type of data are we talking about? You know data as from a thematic perspective it is. There's, we know when every company knows on some level there's tremendous value in the data. The challenge is being able to access it and unlock the value. Give us a slice of and capital markets for example. What are some of the types of data services that you provide to your customers? >>So we have all sorts of data. So we obviously source the data from lots of different sources where it's coming from, from exchanges or from the, from the market data sources. And then our customers use that to analyze the data and really running the back testing for, for those data facts. They also commingled our data with alternative data sets as well, as well as they own internal data. So it's all about that, that analytical layer that they can add on top of our day. >>Okay. And estate as a service essentially. Is that right? We do have some data as a service. We also deliver the data to the client. People are interested in accessing data in all sorts of different ways, including increasingly on the cloud. So talk more about your cloud offering, your, your cloud and your title. Cloud architecture. >> So one of the things that we're doing is we have a combination, we're an interesting company in that we both have our own pieces of cloud infrastructure for our own purposes, but also increasingly we need to build and deliver solutions for our customers to be asked to consume data in the cloud. So that means being able to work with them to put it into the cloud that they want it to be going into, to be able to work out how we can keep that data up to date and to do it in a cost effective manner for our clients to be able to get the most out of it. >> How do you deal with the problems >>of data quality? You're getting data from different sources. How do you take care of that? >>So anyways, that's, that's really all our core strength and expertise that we have. We have been doing that for years and years. So again, coming from it from defense sources, we normalize the data on our side, we clean it up. And then so for our customers in a you our own information model, and we have created this app Poona permanent and unique, identify a post per ID. So we map all the datasets so it's very easy for our customers to consume that and then also map it back to they own data and third party data sets. Where does the global security come into play? Because that's a topic and thing that we talk about at every event when you're talking about all these different external data sources, quality. But security is, I imagine fundamental. How do you help deliver that? Absolutely. Obviously from that, from the cloud perspective, that has been a big theme in the, in the public cloud environment and I think we are seeing more and more feedback from our customers that as it comes down to public cloud, I think they are very comfortable actually now with uh, with the privacy and security of, of public cloud. >>So that has been, I think, big change past couple of years. I haven't personally seen those sponsors anymore coming, coming from customers the way that we saw a couple of years ago. >>Oh, one of the interesting things that we're seeing is an increasing move is that our clients want to be able to mix their data with that data. And so increasingly you're seeing interesting solutions coming to market, which allowed them to keep their data where their data is held on their cloud or even on their own premises and mix that with our data. And so we're trying to bring together those solutions where a customer doesn't have to put all of our data with theirs but all of their data with us. But keep that segregation as you say, because that PII data and all of those sorts of things are much more important these days for us to be able to be able to show that is how the data is being segregated and that things are being kept apart in an appropriate way. >>Who's responsible for that? Is that you guys, is it the cloud provider? Is it on customers? So it's a shared responsibility model. Where does, where do you leave off and where does the customer pick up? What do you advise customers in terms of, Hey, here's what we're going to do for you and now you have to be responsible for X. What does that line? >>Well, I quite often defining that service boundary is something that we continue to work on. So historically we've delivered data to clients and so we've had lines going into a client. It's a, um, premises. And then there's an obvious point at the end of that where this was us and that's you. As we get more into the cloud space, we have to define much more clearly what that service boundary is. So again, as we're developing out some of our cloud propositions, that's a key thing that we're working through as to what is it that the client wants to control and what is it that we need to control. >>It's very true, Hannah, I mean 10 years ago you talk to financial services companies and he said, we will never be in the cloud and now they're much more comfortable. Now you guys do this cloud survey each year. W w what are you seeing? I'll share some of our data. I wonder if it matches what, what do you, what are the big trends? >>Sure. Yeah. So we are doing this, it's almost becoming tradition for us to do this quota. They are on a yearly basis. So it's quite interesting to kind of compare the previous service and where we are today. So what we have found out on the survey this year is that the IOT, uh, investment is very much going to public cloud. So I think when we started the cloud survey a couple of years ago, we saw that about 32% of the ID investment went to public cloud. But then for next year that is increasing almost to two 50% so obviously public cloud is definitely here to stay. I think another, another key trend that we saw from the surveys that I think the testing that the companies have been doing, like they are learning more and more and they are really seeing the benefit from Papa now and I will highlight that especially our hedge fund customers, they were highlighting a face or so of course benefits with that, with the cloud. >>So about 92% so that actually when they moved to the cloud and do the project in the cloud environment, it really saves money for them, which is quite interesting. Payers also then at the same time to work many of the customer discussions. Like it can be also a challenge for, especially for large organizations as they move to the cloud environment, that how do you kind of manage that a traditional technology stack and when you move to the public cloud. So it's kind of two sided way there, but I think the general consensus as it comes down to out survey was that many of the organizations, they really saw that big transition that organizations are going for one that it can be very, very big impact for they own own business. So very, very positive message on that part. >>Let's dig into that a little bit more from a transition or we'll use Andy, Jesse's or a transformation. James, I'd love to get your perspective on what has changed in the last few years to see the numbers that Helen talked about. Um, really Hannah, excuse me, going up so significantly as we know that, you know, cloud one compute and storage and um, networking and maybe some data services. But what do you think has fundamentally changed across industries such that public cloud now is much more strategic? >>I think for a lot of firms and particularly in financial services, we spend a lot of time looking at analytics and being able to run those large analytical jobs and be able to scale them. I think that as people have become more comfortable about the data that they can put into the cloud and being able to get access to more data through companies like definitive, being able to run those machine learning jobs. And it was really interesting to see the keynote this morning to see Amazon really putting a lot of effort into democratizing the use of machine learning through Sage maker thought it was very exciting. Um, we think that that is going to be an increasing thing. So as you see in financial services, people are looking for those large workloads. They have really large data sets and so the only way that they can do that and it kind of realistic manner is being able to use public cloud. And then you see them taking a lot of the old traditional systems. And as we're seeing the risk appetite to be able to get onto cloud becoming more, they're going through the same of transformation, which we see many firms having gone through. You know, the developers are insisting that they're getting the best tools so that it can be, have the agility to deliver what their clients want. And again, one of the best ways of doing that is moving onto a public cloud infrastructure that really delivers those tools to >>what are, if you could talk about what you're seeing in terms of adoption of new tech. So I said we share some of our data at the macro, you know, spending slowing down, it's, it's reverting to pre 2018 levels. It's not falling off a cliff, but, but when you look at the spending data from ETR and others, it's slowing down. Financial services is a bellwether. You're seeing less experimentation and sort of more narrowing of their bets to the placing bets on things that they know are going to work. They've been experimenting with digital transformation for the last couple of years and now they're saying, Hey, we're now going to double down on the things that work. We're going to unplug the things, the legacy stuff so we can get rid of some of our technical debt. What are you seeing in terms of the trends of technology adoption for particularly for emerging tech within Fs? >>Yeah, and I think you've touched on this briefly, but I think what we are seeing is that the, when the, when we started co did the discussions with our customers, they all started with the kind of the backend technology I take on rotation at that time. But I in that trend as you say as well, so it's moving very much to the end users and end users. For example, data scientists speaking the analytical tools if they want to go into them. And I think that's a, that's a very big trend that we are seeing. So again, AI, ML analytics in general that you can add on top of the cloud environment and on top of the data, that will be the big thing happening. >>One of the things that Andy Jassy said this morning, James is in sort of these four kind of essentials for transformation to happen and he said the first one is you've got to get senior executive alignment and the second thing he said is has to be this, and I use the word aggressive, aggressive, top down approach. What are some of the changes that you're seeing with respect to, you know, where it comes to maybe what, what, what you said, Hannah, about the emerging technologies and the end users really in the data scientists needing to be able to get their hands wet with all this, but what are you seeing in terms of organizations that you work with? Where is that senior leadership really getting onboard where public cloud is a strategy that is driven top-down? >>Absolutely. I mean increasingly you're seeing that happen is that it really is going to be the top down strategy. There are a number of very large capital markets firms who have come out and said that they're going to adopt varying cloud providers. And increasingly that's because the level of trust has gone up and the level of maturity of the cloud providers. There's also increased. So a few years ago you would speak to the cloud providers and they really wouldn't understand the need to engage with the regulators. Now companies have large teams of people who go out and engage with the regulators and they will partner with the financial institutions to make sure that we're getting the right sort of level of engagement and the right level of permission to do these things. So that means that the senior management are there. And I think that also the senior management, you know, finally are starting to see some of the benefits flow through in terms of a combination of the agility, the different sort of cost controls and the elasticity. >>And if you think about some of the nature of the workloads that financial institution run, you've got a lot of this overnight processing, which still goes on for creating risk reports and all those sorts of things really well suited for elasticity. And in the last few years you've seen trust this massive increase in the regulatory requirement for those things. And certainly the institutions that I've worked with, you end up in a situation where you're saying, well, in order to be able to accommodate just working out what I need to do there, I'd need to build three different data centers clean. Nobody is doing that anymore. You're going to go out, you're going to partner with your cloud provider and they're going to provide you with that capability. That may not be something that you need in the longterm, but it'll be something that will help you work out what it is that you do need. And then you can turn that into a normal world. >>So AWS, AWS obviously is a cloud provider for you. There may be others as well, but you saw some of the announcements today. You mentioned some of the machine learning and AI stuff, Sage maker, you also saw a lot of activity around the data store, you know red shift and separating computer storage. Is that something that you care about is that your customers have to worry about that? Sometimes they ask you for the solution. >>We super care about this. In fact, one of the big things that we're looking at at the moment, and I was really interested in the announcements today, but exactly that is how do we get our data into people's data lakes? As I said, how do we do that in a way where we're making sure that the commitments that we have on digital rights management are being honored and how do we work with cloud providers like Amazon about how we do that. So we have very strong relationships with Amazon. We have very strong relationships with other providers as well. And so we are trying hard to work out what the best solution is because to be honest with you, we have to deliver where our clients want the data to be. So we're working with lots of different providers on this, but these are all really interesting times and this focus on the data and how you get the data into people's data lakes is really interesting to us and something where we're pushing very hard. >>Yeah. And then, and then how you act on it. It's a whole new layer of compute being driven and new workloads that are emerging as a result of that data. It's not just throw it in the data Lake anymore. It's I have to extract insights. Absolutely. Yeah. >>Talk to us about how on that front, how are you helping him? We'll start with you. How are you helping customers, maybe a large enterprise legacy organization actually start to use data for competitive advantage in business differentiation, especially where the enterprise is concerned, where they most likely have competitors that are born in the cloud, that have the agility and the speed and the appetite to take risks. How are you helping customers unlock this data and go, wow, this is a huge advantage in our business. Absolutely. So obviously as, as I said earlier though, because we are a data company, so our customers know know us from that perspective. So they come to us for, for both financial and risk data. That's kind of one >>go to place to get everything. And then we are obviously working very closely with our customers to also offer them new additional datasets. So things like alternative data obviously being one that you again want to go mingle your own data with a third party data with alternative data sets as well. So we, for example, formed a partnership with a company called Patal Finn earlier this year, which has this very nice technology to onboard different alternative data sets. And then we are onboarding those data sets for our customers. Again, combining that with our overall information model. But it's really, again, coming back to that flexible a question that we want to make sure that all our days are, can be served in the environment where our customers are. So whether they are in public cloud, private cloud, where they have their own prem solution, stale, obviously with, especially with a larger institution, they still have those, uh, as well as we, we hosting the offering for them as well as, or it's all about the flexibility that we will be offering. Excellent. >>Well, Hannah, James, thank you for joining David Mead, sharing with our audience who were fitted. It is what you do and really kind of this importance of data as we're in this new NextGen of cloud. We appreciate your time. Thank you so much for day. Volante I'm Lisa Martin. You're watching the queue from day one of our coverage of AWS reinvent 19. Thanks for watching.

>>From New York city. It's the cube covering. Welcome to the new edge brought to you by systems. >>Okay, we'll come back. You're ready. Jeff Frick here with the cube. We're in downtown Manhattan at the top of Goldman Sachs, like 43 stories above the Hudson. It was a really beautiful view a couple hours ago, but the cloud has moved in and that's only appropriate cause it's cloud is a big theme of why we're here today. We're here for the Penn Zando event. It's called welcome to the new edge. They just come out of stealth mode after two and a half years, almost three years, raised a ton of money, got a really rockstar team and we're excited to have the CEO with us today to tell us a little bit about more what's going on. And that's prem Jane and again, the CEO of Penn Sandow prem. Great to see you. Nice to see you too. So everything we did running up to this event before we could get any of the news, we, we, we tried to figure out what was going on and all it kept coming up was NPLS, NPLS, NPLS, which I thought was a technology, which it is, but it's really about the team. Tell us a little bit about the team in which you guys have built prior and, and why you're such a, a well functioning and kind of forward thinking group of people. >>So I think the team is working together. Mario Luca, myself and Sony were working together since 1983 except for Sony. Sony joined us after the first company, which has crescendo, got acquired by Cisco in 1993 and since then four of us are working together. Uh, we have done many, uh, spinnings inside the Cisco and demo was the first one. Then we did, uh, uh, Nova systems, which was the second, then we did recently in CMA. Uh, and then after we left we thought we are going to retire, but we talked about it and we says, you know, there is still transitions happening in the industry and maybe we have few more years to go back to the, you know, industry and, and do something which is very challenging and, and uh, impacting. I think everything which we have done in the past is to create a impact in the industry and make that transition which is occurring very successful, >>which is really hard to do. And, and John Chambers who, who's on the board and spoke earlier today, you know, kind of talked about these 10 year cycles of significant change in our industry and you know, Clayton Christianson innovator's dilemma, it's really easy when you are successful at one of those to kind of sit on your laurels. In fact, it's really, really hard to kill yourself and go on to the next thing you guys have done this time and time and time again. Is there a unique chemistry in the way you guys look forward or you just, you just get bored with what you built and you want to build something new. I mean, what is some of the magic, because even John said, as soon as he heard that you were the team behind it, he was like, sign me up. I don't know what they're building but I don't really care cause I know these people can deliver. >>I think it's very good the, whenever you look at any startup, the most important thing which comes up as the team and you're seeing a lot of startup fails because the team didn't work together or they got their egos into this one. Since we are working for so long, they compliment each other. That's the one thing which is very important. Mario, Luca, myself, they come from engineering backgrounds. Sony comes from marketing, sales, uh, type of background and we all lady in terms of the brain, if you think about is the Mario behind the scene, Luca is really the execution machine and I'm, you can think like as a heart, okay. Putting this thing together. Uh, as a team, we work very complimentary with each other. It does not mean that we agree on everything, right? We disagree. We argue. We basically challenge each other. But one thing good about this particular team is that once we come to a conclusion, we just focus and execute. And team is also known to work with customers all the time. I mean, even when we started Penn Sando, we talked to many customers in the very beginning. They shape up our ideas, they shape up the directions, which is we are going and what transitions are occurring in the industries and all that. That's another thing which is we take customer very seriously in our thought process of building a product. >>So when you were thinking around sitting around the table, deciding whether you guys wanted to do it again, what were the challenges that you saw? What was the kind of the feedback loop that came in that, that started this? The, uh, the gym of the idea >>thing is also is that, uh, we had, we had developed so many different products as you saw today in the launch, eight or nine, uh, billion dollar product line and stuff like that. So we all have a very good system experience what is really needed, what transitions are occurring and stuff like that. When we started this one, we were not really sure what we wanted to do it, but in the last one when we did the, uh, NCMA, we realize that the enterprise thing, which we deliver the ACI solution for the enterprise, the realize that these services was the most complex way of incorporating into that particular architectures. So right from the beginning of interview realized that the, this particular thing is nobody has touched it, nobody thought about it out of the box thinking that how can you make it into a distributed fashion, which has also realized that cloud is going, everything distributed. >>They got away from the centralized appliances. So as the enterprise is now thinking of doing it cloud-like architectures and stuff like that. And the third thing which was really triggered us also, there was a company which is a new Poona which got acquired by Amazon in 2016 and we were looking at it what kinds of things they are doing and we said we can do much better architecturally and next generation, uh, architecture, which can really enable all the other cloud vendors. Some of them are our partners to make sure they can leverage that particular technologies and build the next generation cloud. And that's where this idea of new edge came in because we also saw that the new applications like IOT is five G's and artificial intelligence, machine learning, robotics or drones, you just name it intelligent devices, which is going to get connected. What is the best place to process them is at the edge or also at the backend with the application where the server is running these and that is another edge compute edge, right? >>In that particular sense. So our idea was to develop a product so that it can cover wide segment of the market, enterprise cloud providers, service borders, but focus very narrowly delivering these services into existing architectures. Also people who are building, building the next generation architectures. Right, so it's the distributed services platform or the distributed services architecture. So at its core for people that didn't make it today, what is it? It's basically is a distributed service platforms. The foundation of that is really our custom processor, which is we have designed is highly programmable. It's software defined so that all the protocols, which is typically people hardwired in our case is programmable. It's all programs which is we are writing the language which you selected as before and before extensions. The software stack is the major differentiated thing which is running on the top of this particular processor, which is we have designed in such a way that is hardware agnostics. >>The the, the capabilities which we have built is easily integrated into the existing environment. So if people already have cloud and they want to leverage our technologies, they can really deploy it in the enterprise. We are basically replacing lot of appliances, simplifying the architectures, making sure they can enable the service as they grow model, which is really amazing because right now they had to say firewall goes here, load balancer goes here, these a VPN devices goes there. In our case it's very simple. You put in every server of our technologies and our software stack and our Venice, which is our policy manager, which is sitting outside and it's based upon Kubernete X a architectures is basically a microservices, which is we are running and managing the life cycle of this particular product family and also providing the visibility and uh, uh, accountability in terms of exactly what is going on in that particular network. >>And it's all driven by intent-based architecture, which is policy driven, right? So software defined sitting on software defined Silicon. So you get the benefits of the Silicon, but it's also programmable Silicon, but it's still, you're sitting, you've got a software stack on top of that that manages that cloud and then the form factors as small as a Nick. Yes. So he can stick it in the HP HP server. Yeah. It specifically goes into any PCI slot in any server, uh, in the industry. Yes. It's amazing. Well, first incarnation, but, but, but, but, but that's a really simple implementation, right? Just to get radiation and easy to deploy. Right. And you guys are, you're yourself where involved in security that's involved in managing the storage. It's simple low power, which I thought was a pretty interesting attribute that you defined early on. Clearly thinking about edge and these distributed, uh, things all over the place. >>They're metal programmable. And then the other thing that was talked about a lot today was the observability. Yes. Um, why observability why was that so important? What were you hearing from customers that were really leading you down that path? Yeah, it's important. Uh, you know, surprisingly enough, uh, the visibility is one of the biggest challenge. Most of the data center faces today. A lot of people tried to do multiple different things, but they're never able to do it, uh, in, in the way we are doing it. One is that we don't run anything on the host. Some people have done it right on the train running the agent on the host. Some people have tried to run virtual machines on the those particular environment. In our case there's nothing which is running on the host site. It runs on our card and having end to end that visibility we can provide latency, very accurate latency to the, to the applications which is very important for these customers. >>Also, what is really going on there is the problem in the network. Isolation is another big thing. When something get lost they don't know where it got lost. We can provide that thing. Another important thing that you're doing, which is not being done in the industries. Everything which is we are doing is flow based means if I'm talking to you, there is a flow being set up between you and me and we are monitoring every flow and one of the advantages of our processor is we have four to eight gigabytes of memory, so we can keep these States, have these flows inside, and that gives a tremendous advantage for us to do lots of things, which as you can imagine going forward, we will be delivering it such as, for example, behavior of these flows and things from this point of view, once you understand the behavior of the flow, you can also provide lot of security features because if I'm not talking to you and suddenly I start talking to you and I know that there's something went wrong, right, right. >>And they should be able to look at the behavior analysis and should be able to tell exactly what's going on. You mean we want a real time snapshot of what's really happening instead of a instead of a sample of something that happened a little. No, absolutely. You're absolutely connected. Yeah. Yeah. Um, that's terrific. So you put together to accompany and you immediately went out and talked to a whole bunch of customers. I was amazed at the number of customers and partners that you had here at the launch. Um, was that for validation? Were you testing hypotheses or, or were there some things that the customers were telling you about that maybe you weren't aware of or maybe didn't get the right priority? I think it's all of the above. What you mentioned our, it's in our DNA by the way. You know, we don't design products, we don't design things without talking to customers. >>Validation is very important that we are on the right track because you may try to solve the customer problem, which is not today's problem. Maybe future's problem. Our idea was that then you can develop the product it was set on the shelf. We don't want to do that. We wanted to make sure that, that this is the hard problem customer is facing today. At the same time looking at it, what futuristic in their architecture is understanding the customers, how, what are they doing today, how they're deploying it. The use cases are understanding those very well and making sure that we are designing. Because when we design a seeker, when your designer processor, you know, you cannot design for one year, it has to be a longterm, right? And you need to make sure that we understand the current problems, we understand the future problems and design that in pretty much your spark and you've been in this space forever. >>You're at Cisco before. And so just love to get your take on exponential growth. You know, such an interesting concept that people have a really hard time grasping exponential growth and we're seeing it clearly with data and data flows and ultimately everything's got to go through the network. I mean, when you, when you think back with a little bit of perspective at the incredible increase in the data flow and the amount of data is being stored and the distribution of these, um, applications now out to the edge and store and compute and take action at the edge, you know, what do you think about, how do you, how do you kind of stay on top of that as somebody who kind of sees the feature relatively effectively, how do you try to stay on top of exponential curves? As you know, very valuable data is very important for anybody in any business. >>Whether it's financial, whether it's healthcare, whether it's, and it's becoming even more and more important because of machine learning, artificial intelligence, which is coming in to really process this particular data and predict certain things which is going to happen, right? We wanted to be close to the data and the closest place to be data is where the application is running. That's one place clears closest to the data at the edge is where data is coming in from the IOT devices, from the 5g devices, from the, you know, you know all kinds of appliances which is being classified under IOT devices. We wanted to be, make sure that we are close to the data, doesn't matter where you deploy and we want to be agnostic. Actually our technologies and architectures designed that this boundary is between North, South, East, West is going to go away in future cloud. >>A lot of things which is being done in the backend will be become at the edge like we talked about before. So we are really a journey which is just starting in this particular detectors and you're going to see a lot more innovations coming from us continuously in this particular directions. And again, based upon the feedback which you're going to get from cloud customers with enterprise customers, but they were partners and other system ecosystem partners, which is going to give us a lot of feedback. Great. Well again, thanks for uh, for having us out and congratulations to uh, to you and the team. It must be really fun to pull the covers off. absolutely. It is very historical day for us. This is something we were waiting for two years and nine months to see this particular date, to have our customers come on the stage and talk about our technologies and why they think it's very important. Thank you very much for giving me this opportunity to talk to you. Thank you. Alright, thanks prem. Thanks. He's prem. I'm Jeff. You're watching the cube where it depends. Sandow launch at the top of Goldman Sachs in downtown Manhattan. Thanks for watching. We'll see you next time.