 System Partners, now here are your hosts. John Furrier and Stu Miniman. You guys, Amazon Web Services, re-invent 2016, their annual conference, 32,000 people, record setting, number, I'm John Furrier, Stu Miniman, co-hosting theCUBE for three days of wall-to-wall covers, day two, day one of the conference, our next guest is Kieran Baggispor, who's the CEO and co-founder of Ignia Systems, CUBE alum, hot startup in the, I don't want to say storage area, kind of disrupting storage in a new way. Kieran, great to see you, thanks for coming on theCUBE. Thanks a lot, glad to be here, John. So you're living the dream, the cloud dream, it's not a nightmare for you because you're one of the progressive new ways. I want to get your thoughts on Andy Jash's keynote because he really lays out the new mindset of the cloud. Your startup that you founded with your team is doing something kind of, I will say contrarian, some might say contrarian, but contrarians usually become the big winners. Like Amazon was a contrarian, now they're obviously the winning. So take a minute to explain what you guys are doing. You're funded by Medrano Ventures and NEA New Enterprise Associates. Great backers, smart. You're track record, Iceland, you know the business. Take a minute to describe what you guys are doing. Great, yes, I will. So Ignia Systems was founded to really deliver cloud services to the enterprise data center for data-centric workloads. So what do we mean by that? With cloud services, just like with Amazon, customers don't buy hardware, license software, they do not monitor, manage your infrastructure. They consume it across API and they pay for it by the sort of drip rather than the drink. Similarly, same case with us, but we make that all available within a customer's data center itself and we focus on sort of data-centric, data-heavy workloads. I don't know whether you saw James Hamilton's speech yesterday, but he also talked about the same thing that Meri Meeker talked about earlier this year, which is an overwhelming amount of data generated today, is machine generated and machine consumed and that's growing really rapidly. And our view is the same techniques that have made Amazon so powerful and so valuable are needed out at the edge or on-premise close to where users and machines are generating and using the data. So that's kind of what we do. So very much the cloud model taken out to the enterprise data center. So think of it as hybrid. Kieran, let's talk about storage and where it lives because I think something that many people miss is that cloud typically starts with very compute-heavy types of applications and we know that data's tough to move. I mean, Amazon rolled out a truck to show how they move 100 better rates and not just to show it, this is a new product they had because customers do want to be able to migrate data and that's really tough and takes a lot of time. You mentioned IoT at the edge. They announced kind of query services on your data up in S3. So what are you hearing from customers? You know kind of large data from your previous jobs. Where's the data living? Where's data being created? Where does data need to be worked on and has that play into what you're doing? That's a great question Stu. What we find with customers, especially the ones with large and growing data sets is there is still a challenge of not just how to go store it but how to go process that on the fly. You know, a camera today or a next generation microscope could produce tens of terabytes of data per hour and that is not stuff that you can move across the internet to the cloud. And so the ask and the call from customers is to be able to go ingest that, curate that, process that locally and the cloud still has a very compelling role to play as a distribution mechanism and for a sharing mechanism of that data. I mean, I found it pretty wild that a big part of Andy Jassy's keynote was for the first time they talked about hybrid and acknowledged the fact that it is the cloud and cloud like techniques out in the enterprise data center. So I look at that as hugely validating what we've been talking about which is bringing cloud native pattern dimes into the sort of enterprise data center. Let's talk about that operational model because what you're highlighting and what Jassy pointed out is an operational model now for IT. How are you guys creating value for customers and be specific, is it? Because the on-prem is not going away. We've talked about this before and certainly VMware sees the cloud but also on-prem too. What is the value for customers? Because now this operational model of one of the cloud is there one way. But how do I get cloud inside my data center? The way we do that is very similar to the cloud operating model, right? So we sell customers essentially an annual subscription service and that service is delivered using appliances that are purpose built. Think of it as like Snowball, if you will, that goes into the customer's data center fully managed by our software running in our cloud. So from a customer point of view, it happens to live within their data center but they are consuming it pretty much the same way that they would consume a cloud service. So that's the sort of value. It's the same tool chains, the same programming paradigms that they are used to with, say, a native US but within their data centers at lower latencies, sort of addressing the same things that Andy just brought up, which is you need a truck to go move large amounts of data. Well, I want to also bring up James Hamilton's presentation. You mentioned that yesterday, one of the key points he made was that scaling up for these peak loads like they have on the Fridays there, the Prime Friday spikes, they do instantly elastic is a big deal and we know that. His point though was they would have the provision on bare metal or in the data center months in advance to even rationalize what that peak could be, which still is an unknown number. So the scale point in provisioning is a huge headache for customers so that's why that's relevant. How do you guys answer that claim when you say, hey, I need stuff to be done fast. I don't have time to provision. How do you guys, do you address that at all? And how do you talk to that specific point? Well, we take care of the provisioning and the additional expansion and shrinking of capacity within the customer data center because just like Amazon monitors their infrastructure users in the data center, we do that for our infrastructure within the customer's data center and therefore we can react to go scale up or scale down. But then there's another point to the whole thing which is the interesting thing is the elasticity is much more important for compute as opposed to data. Data just linearly grows. You never throw that stuff away. The things that you captured, the processing is highly elastic and you might want to do some additional processing and burst out and so on. So that's another aspect of hybrid we see with our customers which is, I want my workflow here. I want to be able to burst out to the public cloud for that peak capacity that I don't want to have infrastructure locally for. Okay, Kira, I'm sorry. So James Hamilton's presentation talks a lot about just that hyper scale and that they claim that they've got the most scale and therefore nobody else should do anything because oversimplifying a little bit, but we've got the best price, we've got the whole stack, give you all the solutions. You talk to enterprises, I mean scale means different things for different applications, for what I need to get done and what I have. What does that really mean to you? How does that hybrid piece fit into the whole scale discussion? A lot of what we do is sort of really right on the coattails of the Amazon and the Googles and the Microsoft because everyone has access to the same raw components, hot drives and CPUs and so on and so forth. And then the question is, how do you go assemble those in a form factor that is appropriate for that particular use case? If you're going to go build a data center that's one level of scale, but if you look at a vast majority of applications and enterprises, their scales are much smaller. So we literally look at taking a rack of infrastructure which might have say 40 servers and a couple of switches in sheet metal and shrinking that to a 4U form factor which has got 60 of our nano servers which has got switches and it's got sheet metal. So it's shrinking the whole thing down. The economies of scale are still quite compelling because we use the exact same raw materials from the same suppliers to the cloud guys, right? And the real difference in cost is how things are put together and how they are operationalized in which case we are much more like Amazon than not. The other thing that's really interesting to watch if you look at Amazon's storage move is, storage isn't a silo, they've now got all these services that I can start doing this. How does the enterprise look at that? How does the solution like yours enable us to be able to use our data more? I absolutely think there is a palpable need for and desire for those sort of new paradigms in the enterprise data center too because what you can do with not just storage but with Lambda and with a bunch of other advanced services on top of that, what that really does is allows enterprises and customers to just focus on what is differentiated to them. So this is the whole low code, no code moment if you will, right? Movement and that's a compelling trend. It is something that we've sort of actively embraced. We've got sort of our architecture enables that on day one and that's kind of the way you're going to go build applications now onwards. So will we see Lambda functions calling things on your end? Stay tuned, I think my, yeah, stay tuned. That's a smile, that's a yes. Talk about the drivers in your business because you guys are new, you're a startup but the folks watching, you're making some bets. Big bets, obviously funded by some pretty big venture capitalists out there. What is your big bet? Is it true private cloud is going to emerge on premise? Is the bet that cloud adoption with scalable compute and storage is going to be unmanaged or managed lists or server lists? I mean, what's the big bet? So our bet is the cloud is going to win and I mean the cloud paradigm which means not consuming infrastructure by the drip rather than the drink across APIs, flexibility, agility is going to win. One answer which is very compelling is the public cloud today. We believe that similar patterns will exist on the on-premise world and we believe we are very well positioned to supply that thing and the infrastructure which shrinks would be very traditional infrastructure and software technology stats which has sort of really existed in the enterprise data center for the last 20 years. That will shrink and everything will look kind of similar as in this highly flexible, highly scalable, very easy to go and put things together and you're going to have very similar patterns in both the public cloud and within your data center. Our Wikibon research team is looking at the practitioner side of the market and one of the things they're observing is among a lot of things is that you're seeing AWS teams come together. We're seeing a century was on earlier talking about same dynamic. That's the pattern that we're seeing is these teams are coming together, some handful of people, the pizza box teams as Jeff Bezos calls it, growing into fully functional, bigger teams. Depending upon that progression, what's your advice to practitioners and how do you add value into this momentum of as they scratch their head, go, okay, we're going to go to the cloud? So they know that's the mandate. How do you help them and why should they look at your solution and where do you fit into that? So one of the things customers and partners tell us is we are a great on ramp to the cloud, if you will. Everybody wants to embrace the new programming patterns, the new programming paradigms. And many people have sort of taken that big leap and done the full shift in one step. You know, you've heard Finra, you've heard Capital One, all of these guys talk, but not everyone is that far out there. And so what we sort of become for these folks is a stepping stone. We are on premise. It allows them to get used to it. They start using the same patterns that can get scaled there. They can decide what workflows remain local and why and what go there. And that's kind of our view. We very much live in the hybrid world to burst out to the world, bring it back as appropriate. Kieran, thanks so much for coming on. I think you really appreciate it. We're getting a break, but I do want to ask one personal question. You're back in the entrepreneurial zeal again. You got the startup. You have some capital, but you're not loaded with cash. A good amount to achieve what you need to do. What's it like for you right now? I mean, what are you believing? What's your guiding principles and what's it like to get back on the entrepreneurial trend now again? It's actually quite exhilarating and liberating to be back in a startup environment because it forces you to focus on what is important, what is urgent and important at all points in time. And a guiding principle for us is less is more. Let's be driven by customers and do what is required there and then slowly extend that out. And actually being a startup and not having infinite money to throw like large legacy players would, frees you from trying to do too many things and focus on only what is important. And that's really key to success. And how are you making the decisions as an executive, like product-wise? Is it more agile as you guys doubling down? Very, very agile. We can move very quickly. Since we are delivering this service, we are continuously updating infrastructure just like Amazon does within their data center. So we can turn around very, very quickly. So I'm very impressed the fact that the Amazon roads out a thousand new features this year. But I can see how that is possible at scale and that's kind of what we're doing. At Isilon, you were very successful in scaling up that generation of web scale. We saw it with Facebook and the apples of the world. What's different now than then, just in the short years between the web scalers dominating to now full multi-cloud, hybrid cloud cloud? In your mind, what's different about the landscape out there? Share your thoughts. I think there's a couple of things. One of them is Isilon was incredible, was a very useful infrastructure with something that was easy to deploy, but it was still that something you built, you managed, you sort of owned, if you will. The big transition is away from that, from build to consume and not worry about that infrastructure at all. And that is not something that you can retrofit into an existing architecture. You have to start from scratch to go do that. So that's the sort of biggest number one. The second one is just the scale is bigger. You heard Andy Jassy say, talk about the exabyte moving problem and he commented on the fact that exabytes are not all that rare. And he's true because you go back 10 years ago, maybe four companies had an exabyte problem. It's now a lot more than that. And so the scale is two to three orders of magnitude larger than when Isilon was growing up. Scales of table stakes and consumption of infrastructure. That's a DevOps ethos gone mainstream. Your Honor, thanks so much for sharing your live here in Las Vegas for Amazon reinventing. I'm John Furrier, Stu Miniman. We'll be back with more live coverage, three days of wall-to-wall coverage. The Cube, we'll be right back.