 Live from San Jose, California, in the heart of Silicon Valley, it's theCUBE. Covering Hadoop Summit 2016, brought to you by Hortonworks. Now, here are your hosts, John Furrier and George Gilbert. Okay, welcome back everyone. We are here live in Silicon Valley in San Jose for Hadoop Summit 2016. This is SiliconANGLE Media's theCUBE. It's our flagship program. We go out to the events and extract the signal from the noise. I'm John Furrier, my co-host George Gilbert, big data analyst at wikibond.com, part of SiliconANGLE Media. Our next guest is Praveen Kanakarya, CEO of IAPITUS Technologies and IAJ and our VP of Product Management and Marketing at Kevos Insights. All one big happy family. Guys, welcome back to theCUBE. Great to see you. We just talked last year. Welcome back. Thank you. Good to be here. So I'd love to talk to you guys because you're out in front of customers doing some all this great work and you've got the product side here. You guys are on the front line. So the first question I really want to dig into is, and we talked last year, it's a year later, what's changed? Is Hadoop moving the needle? Is it inch by inch? And then hopefully the long gain or is it gaining at all? Praveen, your thoughts on customers with their sentiment towards the Hadoop and the Hadoop ecosystem? So there's a small minority of haves and these are the people who have really recognized and seized the opportunity. And they've really benefited. We have our largest customer and I talked about them last year as well. They've got 130 use cases in production. They're rolling out the third version of their data lake with full, massively sophisticated governance infrastructure which the industry doesn't offer. They've built it themselves, stitched it themselves. They file patents in an auto technology company and then I see a lot of other large companies where there's still waiting going on. They're waiting for each other inside. And the opportunity, opportunity, I think it's a culture question. It's the deluge of offerings has further contributed to this paralysis, if you will. But Spark is booming. Cloud is on fire. We've seen the cloud guys are winning. Amazon continues to be the gift that keeps on giving to the industry. Their revenues are up. They're getting more and more services. Seeing Google, Azure, those clouds are developing nicely. So cloud and with Docker containers you're seeing huge application development driving significant accelerated multiple, multiple X multiple of benefits. Much, much faster from time to value. So what we've seen is you're talking about adoption of big data and adoption of work. You see a lot of companies can embark on that strategy and get to a certain point and then things languish. And what we see as some of the biggest inhibitors to getting them to the next step is getting that data usable and useful and accessible by the business users. Unless there's adoption by the business users, the cloud strategy or the big data strategy is not going to be successful. So that's really where we come in. We enable a business user to become useful on the cloud or on the big data platform. Because once you've moved your data there, the business user needs to be able to interact with it and explore it without having to go through a learning curve or wait for the data. So the inhibitor is really the complexity, the lack of responsiveness, the slowness of the environment, right? By bringing Kyvo's insights into the solution, what happens is for a business user, now he can interact with that data using his tools that he's always been using, Excel or Tableau or MicroStrategy or whatever he's most comfortable with. And now he can interact with data at any scale and get the response right away. So I got to ask you guys a question. This brings up a good point. We hear this all the time. And I've been trying to piece it together on every cube show I go on. Infrastructure as code really was the DevOps ethos. Make application developers program the infrastructure. So they don't have to do the provisioning, open up ports, do the load balancing, all the networking stuff, paying the butt for them. All that operational stuff that was a different team. Now DevOps has become mainstream. We see Docker containers, Ops is changing. So that game is one. But the same concept could be applied to data. There is in a way a data as code kind of concept going on where I just want the data access as a fabric or just as a layer. Are you guys seeing that trend? I mean, it might be a little early, but I'm kind of trying to tease it out because the same developer challenge is there. Hey, I need data. But I don't even want to be a data jockey to do all the data. Yeah, so big data was initially started off just being for the data scientist. Somebody you could dive in there and write code and develop algorithms to analyze that data. But you need to get it beyond that point where a business user, they can access it and get value out of it without having to go to an IT person and say, okay, this is the report I want and come back to me in a couple of days when you've got the report created. You need to get out of that mode. It's got to become more self-service, right? So the blog posted, I'm thinking about writing in my head right now as we're talking. Praveena, I want to get your thoughts on it so we can probably write this right now. Headline is the silent pivot. And my theme here is, is this industry going through kind of a silent pivot? I mean, we've heard Arun Murthy say in theCUBE, the ecosystem's no longer just about to do. All these guys out there combined revenues aren't that massive, so it's still relevant. So of course it's relevant. But the question is, they're all trying to find their next thing. So is there a silent pivot going on where they're quietly not pivoting in the sense of, well, we're out of money, I got to pivot to a new thing, kind of like, or is this a shift in strategy how do you see the industry? I mean, is it pivoting? Is it turning? Is it adjusting its course? What's your thoughts on the original Hadoop ecosystem where it was just a few years ago and then where it's going? So one thing Ajay said is allowing data to be consumed very easily and naturally by your business lines, that's been a challenge. And even with the Hadoop ecosystem, that's been a challenge. Security's still broken. You go look at some of the major Hadoop vendors and their security infrastructure is still in data, not to be used in production, those disclaimers. And everybody's in production with that stuff. So that's been a challenge. But I think if one can solve it, and like we have solved at Kivos, is allow data to be quickly and easily accessed by people, by the analysts sitting in various corners, various work groups, and if they can do things very quickly, because then the power is immense. You brought all your corporate data in one place, this opportunity did not exist earlier, and you got all the unstructured data in place. And if you have the governance structure, security and governance structure to make this data consumable by anybody and everybody. So the people behind us this year, going back maybe just a few years ago, we saw just a few years ago, venture-backed companies, we're data wranglers as a service, all these kind of companies. The ones that are left standing and these new ones have kind of appeared, and there's some big players here, you know, Microsoft, HP and others, IBM. Are they pivoting? I mean, are they the same? I mean, do you see them pivoting? Or is there a pivot? Am I off base on that? So another pivot is the clouds turning into the big data lake that this industry, in this room, has been aspiring for. So what do you mean by that? Amazon S3 is a place where people are just putting all their data. And we're seeing this across a lot of enterprises where, all right, let's get all our data into S3. We cannot figure out our internal infrastructure. Let's bypass it, get all the data in one place, and then we'll figure out how to consume it from there. But at least it's in one place. That getting all the data in one place alone is a massive, massive challenge for large enterprise. Is that Hadoop or just data in general? Data in general. And I don't think, I think people have become risen to the level of maturity, where they don't think it necessarily has to be Hadoop. They'll pick and choose, or certain pieces from this stack will show up there. We're being asked to support Spark applications on S3 directly. So this begs the next question. If that's the case, if people are saying, okay, hey, support S3 and have all this Amazon, for instance, that begs the question that tooling should be open. So in essence, the tooling or analytics is pick your tool, market. So that's what our customers are asking for. Very large financial industry, customers deploying us on premise right now, but they know they're going to go to S3 and Amazon in a few months. So they want to make sure that the solution they deploy, it makes no difference that they can deploy the same solution as they move to the cloud. Are they saying, well, maybe someday they're actually asking you to write in the contract. No, this is in the contract. You know, they need- As in SLA. Yeah, yeah. But follow this here. And the POC happened on both infrastructure, internally and on the cloud. So customers are writing into the contract with SLA to have the ability to and or move data, all the data to Amazon, S3. Yes. And the impact to the customer is what? Our system working flawlessly there as well. That's not great. Well, that assumes you're building some sort of intelligence layer in front. So, okay, so why is the customer doing that? What's the reasons? What's the impact of them? Costs, agility, options for the future, multitude of tools, all the above. I think agility is big. Cost is the second one. And I think time to market. Time to market, in my opinion, is the biggest. So, are they picking a tool set or are they just going to have their own, you know, bring your own tool to work kind of thing going on? Yeah. P-Y-O-T. Right, so they're building out their stack of tools right for the infrastructure, right? And as they look at, they're actually looking at five years out, you know, the 2020 architecture and they've already kind of started mapping in the layers. And they're positioning us as kind of the BI consumption layer of that infrastructure, right? So which is great for us to be, you know, considered in those terms. So, you know, for us, I think it makes a lot of sense to be, you know, be able to support different environments and be a solution that, you know, both their business users benefit from and their IT guys are not scared of, right? So they can, as you've got the solution being deployed, not just by a few users that are experts on Hadoop, but by potentially thousands of users in the enterprise, now you've got to be able to take care of, you know, can you scale to deal with thousands of users? Do you have the security to deal with, you know, allowing anybody to access your data? So we have to take care of all of that, you know, the scalability as well as, you know, providing that level of security assurance that, you know, a sales guy from that region can only see his region's data and nothing else, right? So having that fine-grained control over their data. So you don't have a physical division of data marks, you know, where companies had 20, 50, 100, 200 data marks. Now, you only have one data mark. There's only one place where all the enterprise data. What about the use case where they might have data out in the network or other networks? Could be IoT, so there might be a little, I won't say little data marks, but like, okay, centralize the data just so you can build your interface for tools. But what about the use case they have other data? Are they going to move it across the network or do you see kind of a tiered approach or? What's your thoughts? You'll see all of the bow, tiered approach, you know, some degree of federation and also moving of some data, process data, being moved to the central data lake, so on. Sorry. Yeah. A key to all something, I actually wanted to ask something that John asked, which was HDFS was this repository for Hadoop-based processing engines. But it sounds like customers are saying, I put all my data in S3 as the physical repository for any engine, not just Hadoop. And if I want to do some transaction processing, I might move it elsewhere. If I want to put it in a fancy analytic pipeline that includes non-Hadoop engines, I'll put it there. Is that what it's becoming, like the bottom most data layer? So we're seeing both, you know, we have one large customer whose system of record is now HDFS, which is a huge elevation for HDFS. In their context. And then we're seeing others, you know, one other big one with that Ajay pointed to. And they are actually moving everything into S3. So now with the data lake, the data in S3 will be a system of record, I don't think so. But definitely that's the meeting place for data coming from all directions. Look, I got to ask you guys with our limited time we have left, I want to get down to the customer conversations. You know, I always ask this question. I want to find out what's trending in the customer conversations you're having. So whether you want to visualize a word cloud or stack rank set of conversations, what are the top three or four, five conversations you're having with customers, material conversations that you think would be notable to share around one, their needs and what you guys are doing with them. Just in general, that might be interesting for folks to know about. So, you know, there's a variety of things. You know, people are looking for, you know, maturity and something that they can depend on in terms of their deployment. That is a key component of the decision-making process. That is something that we can rely on for our enterprise. You know, so it's kind of moved away from, you know, Hadoop being, it's new in the game, it's early on, we've put up with whatever, you know, kinks it's got. Now we're looking for something solid that they can deploy on, right? So that that's- They want reliability, they want rock-side enterprise creating. The traditional enterprise, you know, capability. And how about application development? Any action going on there? Certainly, Docker containers has been very popular. It brings more interoperability. Is that impacting any of your conversations? So the flexibility, right? The flexibility to be able to provision, you know, resources as you need it and then, you know, remove them when you don't need it. You know, that's kind of- It basically accelerates the DevOps. Absolutely, you know. You don't have to wait for, you know, Hadoop cluster to be provisioned. You know, when I was at Yahoo, it would take three months to get a new cluster provision. Now you can do it in minutes, right? Exactly. Okay, remember those good old days? Racking his deck. The guys were out at lunch. Hey, where's the pager? What's his pager number? Those are glory days. And security is a third one, a big one. It's not very sexy, but- No, it's relevant. It's very relevant and with increased usage in- You guys seeing ransomware out there, a lot of ransomware and customer base, is that mostly niche industries like healthcare? No, we're not seeing- You're not seeing much? No. What about the DevOps winning concepts? How mainstream is DevOps now in your mind? And what's the impact to operations? Because ops and dev would be always at odds with DevOps kind of winning the mind share of Agile. That's putting pressure on ops. What's your take on the level of progress with DevOps going mainstream? We're believers because it just produces agility and time to market. One's not waiting on the other. There's no finger pointing, so we just love it. Okay, final worry. What's going on with the company? Here's the quick update. Well, we just introduced a new version of the product. Bigger, better, faster, but also more secure, enterprise ready. And we've got some really exciting customer wins that you're going to be hearing more about. Any updates from you? No, just piggybacking on what Ajay said. I think his customer wins are turning into requests for investments and I'm trying to deal with them, so. Very exciting. Stay private, stay self-funded. Like theCUBE, self-funded is a good thing, but you're across the threshold, so congratulations. Thanks for your insight, guys. Really thanks for sharing the customer perspective and your success and really where all the action is out in the marketplace. We are here in theCUBE breaking down all the action kind of with relevances, enterprise grade. Here in theCUBE, Ajay and Praveen breaking it down for us. I'm John Furrier with George Gilbert. Thanks for watching. We'll be right back with more. You're watching theCUBE.