(electronic music) >> Host: Live from Midtown Manhattan it's The Cube! Covering Big Data New York City 2017, brought to you by Silicon Angle Media, and its Ecosystem sponsors. >> Okay, welcome back everyone. We are here live, The Cube in New York City for Big Data NYC, this is our fifth year, doing our own event, not with O'Reilly or Cloud Era at Strata Data, which as Hadoop World, Strata Conference, Strata Hadoop, now called Strata Data, probably called Strata AI next year, we're The Cube every year, bringing you all the great data, and what's going on. Entrepreneurs, VCs, thought leaders, we interview them and bring that to you. I'm John Furrier with our next guest, Greg Sands, who's the managing director and founder of Costa Nova ventures in Palo Alto, started out as an entrepreneur himself, then single shingle out there, now he's a big VC firm on a third fund. >> On the third fund. >> Third fund. How much in that fund? >> 175 million dollar fund. >> So now you're a big firm now, congratulations, and really great to see your success. >> Thanks very much. I mean, we're still very much an early stage boutique focused on companies that change the way the world does business, but it is the case that we have a bigger team and a bigger fund, to go do the same thing. >> Well you've been great to work with, I've been following you, we've known each other for a while, watched you left Sir Hill and start Costanova, but what's interesting is that, I can kind of joke and kid you, the VC inside joke about being a big firm, because I know you want to be small, and like to be small, help entrepreneurs, that's your thing. But it's really not a big firm, it's a few partners, but a lot of people helping companies, that's your ethos, that's what you're all about at your firm. Take a minute to just share with the folks the kinds of things you do and how you get involved in companies, you're hands on, you roll up your sleeves. You get out of the way at the right time, you help when you can, share your ethos. >> Yeah, absolutely so the way we think of it is, combining the craft of old school venture capital, with a modern operating team, and so since most founder these days are product-oriented, our job is to think like product people, not think like investors. So we think like product people, we do product level analysis, we do customer discovery, we do, we go ride along on sales calls when we're making investment decisions. And then we do the things that great venture capitalists have done for years, and so for example, at Alatian, who I know has been on the show today, we were able to incubate them in our office for a year, I had many conversations with Sathien after he'd sold the first two or three customers. Okay, who's the next person we hire? Who isn't a founder? Who's going to go out and sell? What does that person look like? Do you go straight to a VP? Or do you hire an individual contributor? Do you hire someone for domain, or do you hire someone for talent? And that's the thing that we love doing. Now we've actually built out an operating team so marketing partner, Martino Alcenco, and Jim Wilson as a sales partner, to really help turn that into a program, so that they can, we can take these founders who find product market fit, and say, how do we help you build the right sales process and marketing process, sales team and marketing team, for your company, your customer, your product? >> Well it's interesting since you mention old school venture capital, I'll get into some of the dynamics that are going on in Silicon valley, but it's important to bring that forward, because now with cloud you can get to critical mass on the fly wheel, on economics, you can see the visibility faster now. >> Greg: Absolutely. >> So the game of the old school venture capitalist is all the same, how do you get to cruising altitude, whatever metaphor you want to use, the key was getting there, and sometimes it took a couple of rounds, but now you can get these companies with five million, maybe $10 million funding, they can have unit economics visibility, scales insight, then the scale game comes in, so that seems to be the secret trick right now in venture is, don't overspend, keep the valuation in range and allows you to look for multiple exits potentially, or growth. Talk about that dynamic, because this is like, I call it the hour glass. You get through the hour glass, everyone's down here, but if you can sneak through and get the visibility on the economics, then you grow quickly. >> Absolutely. I mean, it's exactly right an I haven't heard the hour glass metaphor before but I like it. You want to basically get through the narrows of product market fit and the beginnings of scalable sales and marketing. You don't need to know all the answers, but you can do that in a capital-efficient way, building really solid foundations for future explosive growth, look, everybody loves fast growth and big markets, and being grown into. But the number of people who basically don't build those foundations and then say, go big or go home! And they take a ton of money, and they go spend all the money, doing things that just fundamentally don't work, and they blow themselves up. >> Well this is the hourglass problem. You have, once you get through that unique economics, then you have true scale, and value will increase. Everybody wins there so it's about getting through that, and you can get through it fast with good mentoring, but here's the challenge that entrepreneurs fall into the trap. I call it the, I think I made it trap. And what happens is they think they're on the other side of the hourglass, but they still haven't even gone through the straight and narrow yet, and they don't know it. And what they do is they over fund and implode. That seems to be a major trap I see a lot of entrepreneurs fall into, while I got a 50 million pre on my B round, or some monster valuation, and they get way too much cash, and they're behaving as if they're scaling, and they haven't even nailed it yet. >> Well, I think that's right. So there's certainly, there are stages of product market fit, and so I think people hit that first stage, and they say, oh I've got it. And they try to explode out of the gates. And we, in fact I know one good example of somebody saying, hey, by the way, we're doing great in field sales, and our investors want us to go really fast, so we are going to go inside and we, my job was to hire 50 inside people, without ever having tried it. And so we always preach crawl, walk, run, right? Hire a couple, see how it works. Right, in a new channel. Or a new category, or an adjacent space, and I think that it's helpful to have an investor who has seen the whole picture to say, yeah, I know it looks like light at the end of the tunnel, but see how it's a relatively small dot? You still got to go a little farther, and then the other thing I say is, look, don't build your company to feed your venture capitalist ego. Right? People do these big rounds of big valuations, and the big dog investors say, go, go, go! But, you're the CEO. Your job is analyze the data. >> John: You can find during the day (laughs). >> And say, you know, given what we know, how fast should we go? Which investments should we make? And you've got to own that. And I think sometimes our job is just to be the pulling guard and clear space for the CEO to make good decisions. >> So you know I'm a big fan, so my bias is pretty much out there, love what you guys are doing. Tim Carr is a Pivot North doing the same thing. Really adding value, getting down and dirty, but the question that entrepreneurs always ask me and talk privately, not about you, but in general, I don't want the VC to get in the way. I want them, I don't want them to preach to me, I don't want too many know-it-alls on my board, I want added value, but again, I don't want the preaching, I don't want them to get in the way, 'cause that's the fear. I'm not saying the same about VCs in general, but that's kind of the mentality of an entrepreneur. I want someone who's going to help me, be in the boat with me, but not be in my way. How do you address that concern to the founders who think, not think like that, but might have a fear. >> Well, by the way, I think it's a legitimate fear, and I think it actually is uncorrelated with added value, right? I think the idea that the board has certain responsibilities, and management has certain responsibilities, is incredibly important. And I think, I can speak for myself in saying, I'm quite conscious of not crossing that line, I think you talk. >> John: You got to build a return, that's the thing. >> But ultimately I would say to an entrepreneur, I'd just say, hey look, call references. And by the way, here are 30 names and phone numbers, and call any one of them, because I think that people who are, so a venture capital know-it-all, in the board room, telling CEOs what to do, destroys value. It's sand in the gears, and it's bad for the company. >> Absolutely, I agree 100% >> And some of my, when I talk about being a pulling guard for the CEO, that's what I'm talking about, which is blocking people who are destructive. >> And rolling the block for a touchdown, kind of use the metaphor. Adding value, that's the key, and that's why I wanted to get that out there because most guys don't get that nuance, and entrepreneurs, especially the younger ones. So it's good and important. Okay, let's talk about culture, obviously in Silicon Valley, I get, reading this morning in the Wymo guy, and they're writing it, that's the Silicon Valley, that's not crazy, there's a lot of great people in Silicon Valley, you're one of them. The culture's certainly an innovative culture, there's been some things in the press, inclusion and diversity, obviously is super important. This whole brogrammer thing that's been kind of kicked around. How are you dealing with all that? Because, you know, this is a cultural shift, but I think it's being made out more than it really is, but there's still our core issues, your thoughts on the whole inclusion and diversity, and this whole brogrammer blowback thing. >> Yeah, well so I think, so first of all, really important issues, glad we're talking about them, and we all need to get better. And to me the question for us has been, what role do we play? And because I would say it is a relatively small subset of the tech industry, and the venture capital industry. At the same time the behavior of that has become public is appalling. It's appalling and totally unacceptable, and so the question is, okay, how can we be a part of the stand-up part of the ecosystem, and some of which is calling things out when we see them. Though frankly we work with and hang out with people and we don't see them that often, and then part of which is, how do we find a couple of ways to contribute meaningfully? So for example this summer we ran what we called the Costanova Access Fellowship, intentionally, trying to provide first opportunity and venture capital for people who traditionally haven't had as much access. We created an event in the spring called, Seat at the Table, really, particularly around women in the tech industry, and it went so well that we're running it in New York on October 19th, so if you're a woman in tech in New York, we'd love to see you then. And we're just trying to figure-- >> You're doing it in an authentic way though, you're not really doing it from a promotional standpoint. It's legit. >> Yeah, we're just trying to do, you know, pick off a couple of things that we can do, so that we can be on the side of the good guys. >> So I guess what you're saying is just have high integrity, and be part of the solution not part of the problem. >> That's right, and by the way, both of these initiatives were ones that were kicked off in late 2016, so it's not a reaction to things like binary capital, and the problems at uper, both of which are appalling. >> Self-awareness is critical. Let's get back to the nuts and bolts of the real reason why I wanted you to come on, one was to find out how much money you have to spend for the entrepreneurs that are watching. Give us the update on the last fund, so you got a new fund that you just closed, the new fund, fund three. You have your other funds that are still out there, and some funds reserved, which, what's the number amount, how much are you writing checks for? Give the whole thesis. >> Absoluteley. So we're an early stage investor, so we lead series A and seed financing companies that change the way the world does business, so up and down the stack, a business-facing software, data-driven applications. Machine-learning and AI driven applications. >> John: But the filter is changing the way the world works? >> The way, yes, but in particularly the way the world does business. You can think of it as a business-facing software stack. We're not social media investors, it's not what we know, it's not what we're good at. And it includes security and management, and the data stack and-- >> Joe: Enterprise and emerging tech. >> That's right. And the-- >> And every crazy idea in between. >> That's right. (laughs) Absolutely, and so we're participate in or leave seed financings as most typically are half a million to maybe one and a quarter, and we'll lead series A financing, small ones might be two or two and a half million dollars at the outer edge is probably a six million dollar check. We were just opening up in the next couple of days, a thousand square feet of incubation space at world headquarters at Palo Alto. >> John: Nice. >> So Alation, Acme Ticketing and Zen IQ are companies that we invested in. >> Joe: What location is this going to be at? >> That's, near the Fills in downtown Palo Alto, 164 staff, and those three companies are ones where we effectively invested at formation and incubated it for a year, we love doing that. >> At the hangout at Philsmore and get the data. And so you got some funds, what else do you have going on? 175 million? >> So one was a $100 million fund, and then fund two was $135 million fund, and the last investment of fund two which we announced about three weeks ago was called Roadster, so it's ecommerce enablement for the modern dealerships. So Omnichannel and Mobile First infrastructure for auto-dealers. We have already closed, and had the first board meeting for the first new investment of fund three, which isn't yet announced, but in the land of computer vision and deep learning, so a couple of the subjects that we care deeply about, and spend a lot of time thinking about. >> And the average check size for the A round again, seed and A, what do you know about the? The lowest and highest? >> The average for the seed is half a million to one and a quarter, and probably average for a series A is four or five. >> And you'll lead As. >> And we will lead As. >> Okay great. What's the coolest thing you're working on right now that gets you excited? It doesn't have to be a portfolio company, but the research you're doing, thing, tires you're kicking, in subjects, or domains? >> You know, so honestly, one of the great benefits of the venture capital business is that I get up and my neurons are firing right away every day. And I do think that for example, one of the things that we love is is all of the adulant infrastructure and so we've got our friends at Victor Ops that are in the middle of that space, and the thinking about how the modern programmer works, how everybody-- >> Joe: Is security on your radar? >> Security is very much on our radar, in fact, someone who you should have on your show is Asheesh Guptar, and Casey Ella, so she's just joined Bug Crowd as the CEO and Casey moves over to CTO, and the word Bug Bounty was just entered into the Oxford Dictionary for the first time last week, so that to me is the ultimate in category creation. So security and dev ops tools are among the things that we really like. >> And bounties will become the norm as more and more decentralized apps hit the scene. Are you doing anything on decentralized applications? I'm not saying Blockchain in particular, but Blockchain like apps, distributing computing you're well versed on. >> That's right, well we-- >> Blockchain will have an impact in your area. >> Blockchain will have an impact, we just spent an hour talking about it in the context our off site in Decosona Lodge in Pascadero, it felt like it was important that we go there. And digging into it. I think actually the edge computing is actually more actionable for us right now, given the things that we're, given the things that we're interested in, and we're doing and they, it is just fascinating how compute centralizes and then decentralizes, centralizes and then decentralizes again, and I do think that there are a set of things that are fascinating about what your process at the edge, and what you send back to the core. >> As Pet Gelson here said in the QU, if you're not out in front of that next wave, you're driftwood, a lot of big waves coming in, you've seen a lot of waves, you were part of one that changed the world, Netscape browser, or the business plan for that first project manager, congratulations. Now you're at a whole nother generation. You ready? (laughs) >> Absolutely, I'm totally ready, I'm ready to go. >> Greg Sands here in The Cube in New York City, part of Big Data NYC, more live coverage with The Cube after this short break, thanks for watching. (electronic jingle) (inspiring electronic music)

>> Announcer: Live from Midtown Manhattan. It's the CUBE. Covering Big Data New York City 2017. Brought to you by Silicon Angle Media and its ecosystem sponsor. (electronic music) >> Hey, welcome back everyone, live here in New York. This is the CUBE here in Manhattan for Big Data NYC's three days of coverage. We're one day three, things are starting to settle in, starting to see the patterns out there. I'll say it's Big Data week here, in conjunction with Hadoop World, formerly known as Strata Conference, Strata-Hadoop, Strata-Data, soon to be Strata-AI, soon to be Strata-IOT. Big Data, Mike Maccaux who's the Global Big Data Practice Lead at Dell EMC. We've been in this world now for multiple years and, well, what a riot it's been. >> Yeah, it has. It's been really interesting as the organizations have gone from their legacy systems, they have been modernizing. And we've sort of seen Big Data 1.0 a couple years ago. Big Data 2.0 and now we're moving on sort of the what's next? >> Yeah. >> And it's interesting because the Big Data space has really lagged the application space. You talk about microservices-based applications, and deploying in the cloud and stateless things. The data technologies and the data space has not quite caught up. The technology's there, but the thinking around it, and the deployment of those, it seems to be a slower, more methodical process. And so what we're seeing in a lot of enterprises is that the ones that got in early, have built out capabilities, are now looking for that, how do we get to the next level? How do we provide self-service? How do we enable our data scientists to be more productive within the enterprise, right? If you're a startup, it's easy, right? You're somewhere in the public cloud, you're using cloud based API, it's all fine. But if you're an enterprise, with the inertia of those legacy systems and governance and controls, it's a different problem to solve for. >> Let's just face it. We'll just call a spade a spade. Total cost of ownership was out of control. Hadoop was great, but it was built for something that tried to be something else as it evolved. And that's good also, because we need to decentralize and democratize the incumbent big data warehouse stuff. But let's face it, Hadoop is not the game anymore, it's everything else. >> Right, yep. >> Around it so, we've seen that, that's a couple years old. It's about business value right now. That seems to be the big thing. The separation between the players that can deliver value for the customers. >> Matt: Yep. >> And show a little bit of headroom for future AI things, they've seen that. And have the cloud on premise play. >> Yep. >> Right now, to me, that's the call here. What do you, do you agree? >> I absolutely see it. It's funny, you talk to organizations and they say, "We're going cloud, we're doing cloud." Well what does that mean? Can you even put your data in the cloud? Are you allowed to? How are you going to manage that? How are you going to govern that? How are you going to secure that? So many organizations, once they've asked those questions, they've realized, maybe we should start with the model of cloud on premise. And figure out what works and what doesn't. How do users actually want to self serve? What do we templatize for them? And what do we give them the freedom to do themselves? >> Yeah. >> And they sort of get their sea legs with that, and then we look at sort of a hybrid cloud model. How do we be able to span on premise, off premise, whatever your public cloud is, in a seamless way? Because we don't want to end up with the same thing that we had with mainframes decades ago, where it was, IBM had the best, it was the fastest, it was the most efficient, it was the new paradigm. And then 10 years later, organizations realized they were locked in, there was different technology. The same thing's true if you go cloud native. You're sort of locked in. So how do you be cloud agnostic? >> How do you get locked in a cloud native? You mean with Amazon? >> Or any of them, right? >> Okay. >> So they all have their own APIs that are really good for doing certain things. So Google's TensorFlow happens to be very good. >> Yeah. Amazon EMR. >> But you build applications that are using those native APIS, you're sort of locked. And maybe you want to switch to something else. How do you do that? So the idea is to-- >> That's why Kubernetes is so important, right now. That's a very key workload and orchestration container-based system. >> That's right, so we believe that containerization of workloads that you can define in one place, and deploy anywhere is the path forward, right? Deploy 'em on prem, deploy 'em in a private cloud, public cloud, it doesn't matter the infrastructure. Infrastructure's irrelevant. Just like Hadoop is sort of not that important anymore. >> So let me get your reaction on this. >> Yeah. So Dell EMC, so you guys have actually been a supplier. They've been the leading supplier, and now with Dell EMC across the portfolio of everything. From Dell computers, servers and what not, to storage, EMC's run the table on that for many generations. Yeah, there's people nippin' at your heels like Pure, okay that's fine. >> Sure. It's still storage is storage. You got to store the data somewhere, so storage will always be around. Here's what I heard from a CXO. This is the pattern I hear, but I'll just summarize it in one conversation. And then you can give a reaction to it. John, my life is hell. I have application development investment plan, it's just boot up all these new developers. New dev ops guys. We're going to do open source, I got to build that out. I got that, trying to get dev ops going on. >> Yep. >> That's a huge initiative. I got the security team. I'm unbundling from my IT department, into a new, difference in a reporting to the board. And then I got all this data governance crap underneath here, and then I got IOT over the top, and I still don't know where my security holes are. >> Yep. And you want to sell me what? (Matt laughs) So that's the fear. >> That's right. >> Their plates are full. How do you guys help that scenario? You walk in, actually security's pretty much, important obviously you can see that. But how do you walk into that conversation? >> Yeah, it's sort of stop the madness, right? >> (laughs) That's right. >> And all of that matters-- >> No, but this is all critical. Every room in the house is on fire. >> It is. >> And I got to get my house in order, so your comment to me better not be hype. TensorFlow, don't give me this TensorFlow stuff. >> That's right. >> I want real deal. >> Right, I need, my guys are-- >> I love TensorFlow but, doesn't put the fire out. >> They just want spark, right? I need to speed up my-- >> John: All right, so how do you help me? >> So, what we'd do is, we want to complement and augment their existing capabilities with better ways of scaling their architecture. So let's help them containerize their big data workload so that they can deploy them anywhere. Let's help them define centralized security policies that can be defined once and enforced everywhere, so that now we have a way to automate the deployment of environments. And users can bring their own tools. They can bring their data from outside, but because we have intelligent centralized policies, we can enforce that. And so with our elastic data platform, we are doing that with partners in the industry, Blue Talent and Blue Data, they provide that capability on top of whatever the customer's infrastructure is. >> How important is it to you guys that Dell EMC are partnering. I know Michael Dell talks about it all the time, so I know it's important. But I want to hear your reaction. Down in the trenches, you're in the front lines, providing the value, pulling things together. Partnerships seem to be really important. Explain how you look at that, how you guys do your partners. You mentioned Blue Talent and Blue Data. >> That's right, well I'm in the consulting organization. So we are on the front lines. We are dealing with customers day in and day out. And they want us to help them solve their problems, not put more of our kit in their data centers, on their desktops. And so partnering is really key, and our job is to find where the problems are with our customers, and find the best tool for the best job. The right thing for the right workload. And you know what? If the customer says, "We're moving to Amazon," then Dell EMC might not sell any more compute infrastructure to that customer. They might, we might not, right? But it's our job to help them get there, and by partnering with organizations, we can help that seamless. And that strengthens the relationship, and they're going to purchase-- >> So you're saying that you will put the customer over Dell EMC? >> Well, the customer is number one to Dell EMC. Net promoter score is one of the most important metrics that we have-- >> Just want to make sure get on the record, and that's important, 'cause Amazon, and you know, we saw it in Net App. I've got to say, give Net App credit. They heard from customers early on that Amazon was important. They started building into Amazon support. So people saying, "Are you crazy?" VMware, everyone's saying, "Hey you capitulated "by going to Amazon." Turns out that that was a damn good move. >> That's right. >> For Kelsinger. >> Yep. >> Look at VM World. They're going to own the cloud service provider market as an arms dealer. >> Yep. >> I mean, you would have thought that a year ago, no way. And then when they did the deal, they said, >> We have really smart leadership in the organization. Obviously Michael is a brilliant man. And it sort of trickles on down. It's customer first, solve the customer's problems, build the relationship with them, and there will be other things that come, right? There will be other needs, other workloads. We do happen to have a private cloud solution with Virtustream. Some of these customers need that intermediary step, before they go full public, with a hosted private solution using a Virtustream. >> All right, so what's the, final question, so what's the number one thing you're working on right now with customers? What's the pattern? You got the stack rank, you're requests, your deliverables, where you spend your time. What's the top things you're working on? >> The top thing right now is scaling architectures. So getting organizations past, they've already got their first 20 use cases. They've already got lakes, they got pedabytes in there. How do we enable self service so that we can actually bring that business value back, as you mentioned. Bring that business value back by making those data scientists productive. That's number one. Number two is aligning that to overall strategy. So organizations want to monetize their data, but they don't really know what that means. And so, within a consulting practice, we help our customers define, and put a road map in place, to align that strategy to their goals, the policies, the security, the GDP, or the regulations. You have to marry the business and the technology together. You can't do either one in isolation. Or ultimately, you're not going to be efficient. >> All right, and just your take on Big Data NYC this year. What's going on in Manhattan this year? What's the big trend from your standpoint? That you could take away from this show besides it becoming a sprawl of you know, everyone just promoting their wares. I mean it's a big, hyped show that O'Reilly does, >> It is. >> But in general, what's the takeaway from the signal? >> It was good hearing from customers this year. Customer segments, I hope to see more of that in the future. Not all just vendors showing their wares. Hearing customers actually talk about the pain and the success that they've had. So the Barclay session where they went up and they talked about their entire journey. It was a packed room, standing room only. They described their journey. And I saw other banks walk up to them and say, "We're feeling the same thing." And this is a highly competitive financial services space. >> Yeah, we had Packsotta's customer on Standard Bank. They came off about their journey, and how they're wrangling automating. Automating's the big thing. Machine learning, automation, no doubt. If people aren't looking at that, they're dead in my mind. I mean, that's what I'm seeing. >> That's right. And you have to get your house in order before you can start doing the fancy gardening. >> John: Yeah. >> And organizations aspire to do the gardening, right? >> I couldn't agree more. You got to be able to drive the car, you got to know how to drive the car if you want to actually play in this game. But it's a good example, the house. Got to get the house in order. Rooms are on fire (laughs) right? Put the fires out, retrench. That's why private cloud's kicking ass right now. I'm telling you right now. Wikibon nailed it in their true private cloud survey. No other firm nailed this. They nailed it, and it went viral. And that is, private cloud is working and growing faster than some areas because the fact of the matter is, there's some bursting through the clouds, and great use cases in the cloud. But, >> Yep. >> People have to get the ops right on premise. >> Matt: That's right, yep. >> I'm not saying on premise is going to be the future. >> Not forever. >> I'm just saying that the stack and rack operational model is going cloud model. >> Yes. >> John: That's absolutely happening, that's growing. You agree? >> Absolutely, we completely, we see that pattern over and over and over again. And it's the Goldilocks problem. There's the organizations that say, "We're never going to go cloud." There's the organizations that say, "We're going to go full cloud." For big data workloads, I think there's an intermediary for the next couple years, while we figure out operating pulse. >> This evolution, what's fun about the market right now, and it's clear to me that, people who try to get a spot too early, there's too many diseconomies of scale. >> Yep. >> Let the evolution, Kubernetes looking good off the tee right now. Docker containers and containerization in general's happened. >> Yep. >> Happening, dev ops is going mainstream. >> Yep. >> So that's going to develop. While that's developing, you get your house in order, and certainly go to the cloud for bursting, and other green field opportunities. >> Sure. >> No doubt. >> But wait until everything's teed up. >> That's right, the right workload in the right place. >> I mean Amazon's got thousands of enterprises using the cloud. >> Yeah, absolutely. >> It's not like people aren't using the cloud. >> No, they're, yeah. >> It's not 100% yet. (laughs) >> And what's the workload, right? What data can you put there? Do you know what data you're putting there? How do you secure that? And how do you do that in a repeatable way. Yeah, and you think cloud's driving the big data market right now. That's what I was saying earlier. I was saying, I think that the cloud is the unsubtext of this show. >> It's enabling. I don't know if it's driving, but it's the enabling factor. It allows for that scale and speed. >> It accelerates. >> Yeah. >> It accelerates... >> That's a better word, accelerates. >> Accelerates that horizontally scalable. Mike, thanks for coming on the CUBE. Really appreciate it. More live action we're going to have some partners on with you guys. Next, stay with us. Live in Manhattan, this is the CUBE. (electronic music)

>> Narrator: Live from Midtown Manhattan it's theCUBE! Covering Big Data, New York City 2017. Brought to you by SiliconANGLE Media and its ecosystem sponsor. >> Okay, welcome back everyone. Live here in New York this is theCUBE's coverage of Big Data NYC, our event. We've been doing it for five years, it's our event in conjunction with Strata Data, which is the O'Reilly Media that we run, it's a separate event. But we've been covering the Big Data for eight years since 2010, Hadoop World. This is theCUBE. Of course theCUBE is never going to change, they might call it Strata AI next year, whatever trend that they might see. But we're going to keep it theCUBE. This is in New York City, our eighth year of coverage. Guys, welcome to theCUBE. Our next two guests is Andrew Burt, Chief Privacy Officer and Andrew Gillman, Chief Customer Officer and CMO. It's a start-up so you got all these fancy titles, but you're on the A-team from Immuta. Hot start-up. Welcome to theCUBE. Great to see you again. >> Thanks for having us, appreciate it. >> Okay, so you guys are the start-up feature here this week on theCUBE, our little segment here. I think you guys are the hottest start-up that is out there and that people aren't really talking a lot about. So you guys are brand new, you guys have got a really good reputation. Getting a lot of props inside the community. Especially in the people who know data, data science, and know some of the intelligence organizations. But respectful people like Dan Hutchin says you guys are rockstars and doing great. So why all the buzz inside the community? Now you guys are just starting to go to the market? What's the update on the company? >> So great story. Founded in 2014, (mumbles) Investment, it was announced earlier this year. And the team, group of serial entrepreneurs sold their last company CSC, ran the public sector business for them for a while. Really special group of engineers and technologists and data scientists. Headquartered out of D.C. Customer success organization out of Columbus, Ohio, and we're servicing Fortune 100 companies. >> John: So Immuta, I-M-M-U-T-A. >> Immuta.com we just launched the new website earlier this week in preparation for the show. And the easiest way-- >> Immuta, immutable, I mean-- >> Immutable, I'm sure there's a backstory. >> Immutable, yeah. We do not ever touch the raw data. So we're all about managing risk and managing the integrity of the data. And so risk and integrity and security are baked into everything we do. We want our customers to know that their data will be immutable, and that in using us they'll never pose an additional risk to that underlying data. >> I think of blockchain when I think of immutability, like I'm so into blockchaining these dayS as you guys know, I've been totally into it. >> There's no blockchain in their technology. >> I know, but let's get down to why the motivation to enter the market. There's a lot of noisy stuff out there. Why do we need another unified platform? >> The big opportunity that we saw was, organizations had spent basically the past decade refining and upgrading their application infrastructure. But in doing so under the guise of digital transformation. We've really built that organization's people processes to support monolithic applications. Now those applications are moving to the cloud, they're being rearchitected in a microsurfaces architecture. So we have all this data now, how do we manage it for the new application, which we see is really algorithm-centric? The Amazons of the world have proven, how do you compete against anyone? How do you disrupt any industry? That's operationalize your data in a new way. >> Oh, they were developer-centric right? They were very focused on the developer. You guys are saying you're algorithm-centric, meaning the software within the software kind of thing. >> It's really about, we see the future enterprise to compete. You have to build thousands of algorithms. And each one of those algorithms is going to do something very specific, very precise, but faster than any human can do. And so how do you enable an application, excuse me, an algorithm-centric infrastructure to support that? And today, as we go and meet with our customers and other groups, the people, the processes, the data is everywhere. The governance folks who have to control how the data is used, the laws are dynamic. The tooling is complex. So this whole world looks very much like pre-DevOps IT, or pre-cloud IT. It takes on average between four to six months to get a data scientist up and running on a project. >> Let's get into the company. I wanted to just get that gist out, put some context. I see the problem you solve: a lot of algorithms out there, more and more open sources coming up to the scene. With the Linux Foundation, having their new event Rebrand the Open Source summit, shows exponential growth in open source. So no doubt about it, software's going to be new guys coming on, new gals. Tons of software. What is the company positioning? What do you guys do? How many employees? Let's go down by the numbers and then talk about the problem that you solve. >> Okay, cool. So, company. We'll be about 40 people by Q1. Heavy engineering, go to market. We're operating and working with, as I mentioned, Fortune 100 clients. Highly regulated industries. Financial services, healthcare, government, insurance, et cetera. So where you have lots of data that you need to operationalize, that's very sensitive to use. What else? Company positioning. So we are positioned as data management for data science. So the opportunity that we saw, again, managing data for applications is very different than managing data for algorithm development, data sciences. >> John: So you're selling to the CDO, Chief Data Officer? Are you selling to the analytics? >> In a lot of our customers, like in financial services, we're going right into the line of business. We're working with managing directors who are building next generation analytics infrastructure that need to unify and connect the data in a new way that's dynamic. It's not just the data that they have within their organization, they're looking to bring data in from outside. They want to also work collaboratively with governance professionals and lawyers who in financial services, they are, you know, we always jest in the company that different organizations have these cool new tools, like data scientists have all their new tools. And the data owners have flash disks and they have all this. But the governance people still have Microsoft Word. And maybe the newer tools are like Wikis. So now we can get it off of Word and make it shareable. But what we allow them to do is, and what Andrew Burt has really driven, is the ability for you to take internal logic, internal policies, external regulations, and put them into code that becomes dynamically enforceable as you're querying the data, as you're using it, to train algorithms, and to drive, mathematical decision-making in the enterprise. >> Let's jump into some of the privacy. You're the Chief Privacy Officer, which is codeword for you're doing all the governance stuff. And there's a lot of stuff business-wise that's going on around GDPR which is actually relevant. There's a lot of dollars on table for that too, so it's probably good for business. But there's a lot of policy stuff going on. What's going on with you guys in this area? >> So I think policy is really catching up to the world of big data. We've known for a very long time that data is incredibly important. It's the lifeblood of an increasingly large number of organizations, and because data is becoming more important, laws are starting to catch up. I think GDPR is really, it's hot to talk about. I think it is just the beginning of a larger trend. >> People are scared. People are nervous. It's like they don't know, this could be a blank check that they're signing away. The enforcement side is pretty outrageous. >> So I mean-- >> Is that right? I mean people are scared, or do you think? >> I think people are terrified because they know that its important, and they're also terrified because data scientists, and folks in IT have never really had to think very seriously about implementing complex laws. I think GDPR is the first example of laws, forcing technology to basically blend software and law. The only way, I mean one of our theses is, the only way to actually solve for GDPR is to invent laws within the software you're using. And so, we're moving away from this meetings and memos type approach to governing data, which is very slow and can take months, and we need it to happen dynamically. >> This is why I wanted to bring you guys in. Not only, Andrew, we knew each other from another venture, but what got my attention for you guys was really this intersection between law and society and tech. And this is just the beginning. You look at the tell-signs there. Peter Burris who runs research for Wikibon coined the term programming the real world. Life basically. You've got wearables, you've got IOT, this is happening. Self-driving cars. Who decides what side of the street people walk on now? Law and code are coming together. That's algorithm. There'll be more of them. Is there an algorithm for the algorithms? Who teaches the data set, who shares the data set? Wait a minute, I don't want to share my data set because I have a law that says I can't. Who decides all this stuff? >> Exactly. We're starting to enter a world where governments really, really care about that stuff. Just in-- >> In Silicon Valley, that's not in their DNA. You're seeing it all over the front pages of the news, they can't even get it right in inclusion and diversity. How can they work with laws? >> Tension is brewing. In the U.S. our regulatory environment is a little more lax, we want to see innovation happen first and then regulate. But the EU is completely different. Their laws in China and Russia and elsewhere around the world. And it's basically becoming impossible to be a global organization and still take that approach where you can afford to be scared of the law. >> John: I don't know how I feel about this because I get all kinds of rushes of intoxication to fear. Look at what's going on with Bitcoin and Blockchain, underbelly is a whole new counterculture going on around in-immutable data. Anonymous cultures, where they're complete anonymous underbellies going on. >> I think the risk-factors going up, when you mentioned IOTs, so its where you are and your devices and your home. Now think about 23 and Me, Verily, Freenome, where you're digitizing your DNA. We've already started to do that with MRIs and other operations that we've had. You think about now, I'm handing over my DNA to an organization because I want find out my lineage. I want to learn about where I came from. How do I make sure that the derived data off of that digital DNA is used properly? Not just for me, as Andrew, but for my progeny. That introduces some really interesting ethical issues. It's an intersection of this new wave of investment, to your point, like in Silicon Valley, of bringing healthcare into data science, into technology and the intersection. And the underpinning of the whole thing is the data. How do we manage the data, and what do we do-- >> And AI really is the future here. Even though machine-learning is the key part of AI, we just put out an article this morning on SiliconANGLE from Gina Smith, our new writer. Google Brain Chief: AI tops humans in computer vision, and healthcare will never be the same. They talk about little things, like in 2011 you can barely do character recognition of pictures, now you can 100%. Now you take that forward, in Heidelberg, Germany, the event this week we were covering the Heidelberg Laureate Forum, or HLF 2017. All the top scientists were there talking about this specific issue of, this is society blending in with tech. >> Absolutely. >> This societal impact, legal impact, kind of blending. Algorithms are the only thing that are going to scale in this area. This is what you guys are trying to do, right? >> Exactly, that's the interesting thing. When you look at training models and algorithms in AI, right, AI is the new cloud. We're in New York, I'm walking down the street, and there's the algorithm you're writing, and everything is Ernestine Young. Billboards on algorithms, I mean who would have thought, right? An AI. >> John: theCUBE is going to be an AI pretty soon. "Hey, we're AI! "Brought to you by, hey, Siri, do theCUBE interview." >> But the interesting part of the whole AI and the algorithm is you have n number of models. We have lots of data scientists and AI experts. Siri goes off. >> Sorry Siri, didn't mean to do that. >> She's trying to join the conversation. >> Didn't mean to insult you, Siri. But you know, it's applied math by a different name. And you have n number of models, assuming 90% of all algorithms are single linear regression. What ultimately drives the outcome is going to be how you prepare and manage the data. And so when we go back to the governance story. Governance in applications is very different than governance in data science because how we actually dynamically change the data is going to drive the outcome of that algorithm directly. If I'm in Immuta, we connect the data, we connect the data science tools. We allow you to control the data in a unique way. I refer to that as data personalization. It's not just, can I subscribe to the data? It's what does the data look like based on who I am and what those internal and external policies are? Think about this for example, I'm training a model that doesn't mask against race, and doesn't generalize against age. What do you think is going to happen to that model when it goes to start to interact? Either it's delivered as-- >> Well context is critical. And the usability of data, because it's perishable at this point. Data that comes in quick is worth more, but historically the value goes down. But it's worth more when you train the machine. So it's two different issues. >> Exactly. So it's really about longevity of the model. How can we create and train a model that's going to be able to stay in? It's like the new availability, right? That it's going to stay, it's going to be relevant, and it's going to keep us out of jail, and keep us from getting sued as long as possible. >> Well Jeff Dean, I just want to quote one more thing to add context. I want to ask Andrew over here about his view on this. Jeff Dean, the Google Brain Chief behind all of the stuff is saying AI-enabled healthcare. The sector's set to grow at an annual rate of 40% through 2021, when it's expected to hit 6.6 billion spent on AI-enabled healthcare. 6.6 billion. Today it's around 600 million. That's the growth just in AI healthcare impact. Just healthcare. This is going to go from a policy privacy issue, One, healthcare data has been crippled with HIPPA slowing us down. But where is the innovation going to come from? Where's the data going to be in healthcare? And other verticals. This is one vertical. Financial services is crazy too. >> I mean, honestly healthcare is one of the most interesting examples of applied AI, and it's because there's no other realm, at least now, where people are thinking about AI, and the risk is so apparent. If you get a diagnosis and the doctor doesn't understand why it's very apparent. And if they're using a model that has a very low level of transparency, that ends up being really important. I think healthcare is a really fascinating sector to think about. But all of these issues, all of these different types of risks that have been around for a while are starting to become more and more important as AI takes-- >> John: Alright, so I'm going to wrap up here. Give you guys both a chance, and you can't copy each other's answer. So we'll start with you Andrew over here. Explain Immuta in a simple way. Someone who's not in the industry. What do you guys do? And then do a version for someone in the industry. So elevator pitch for someone who's a friend, who's not in the industry, and someone who is. >> So Immuta is a data management platform for data science. And what that actually gives you is, we take the friction out of trying to access data, and trying to control data, and trying to comply with all of the different rules that surround the use of that data. >> John: Great, now do the one for normal people. >> That was the normal pitch. >> Okay! (laughing) I can't wait to hear the one for the insiders. >> And then for the insiders-- >> Just say, "It's magic". >> It's magic. >> We're magic, you know. >> Coming from the infrastructure role, I like to refer to it as a VMWare for data science. We create an abstraction layer than sits between the data and the data science tools, and we'll dynamically enforce policies based on the values of the organization. But also, it drives better outcomes. Because today, the data owners aren't confident that you're going to do with the data what you say you're going to do. So they try to hold it. Like the old server-huggers, the data-huggers. So we allowed them to unlock that and make it universally available. We allow the governance people to get off those memos, that have to be interpreted by IT and enforced, and actually allow them to write code and have it be enforced as the policy mandates. >> And the number one problem you solve is what? >> Accelerate with confidence. We allow the data scientists to go and build models faster by connecting to the data in a way that they're confident that when they deploy their model, that it's going to go into production, and it's going to stay into production for as long as possible. >> And what's the GDPR angle? You've got the legal brain over here, in policy. What's going on with GDPR? How are you guys going to be a solution for that? >> We have the most, I'd say, robust option of policy enforcement on data, I think, available. We make it incredibly easy to comply with GDPR. We actually put together a sample memo that says, "Here's what it looks like to comply with GDPR." It's written from a governance department, sent to the internal data science department. It's about a page and a half long. We actually make that very onerous process-- >> (mumbles) GDPR, you guys know the size of that market? In terms of spend that's going to be coming around the corner? I think it's like the Y2K problem that's actually real. >> Exactly, it feels the same way. And actually Andrew and his team have taken apart the regulation article by article and have actually built-in product features that satisfy that. It's an interesting and unique--- >> John: I think it's really impressive that you guys bring a legal and a policy mind into the product discussion. I think that's something that I think you guys are doing a little bit different than I see anyone out there. You're bringing legal and policy into the software fabric, which is unique, and I think it's going to be the standard in my opinion. Hopefully this is a good trend, hopefully you guys keep in touch. Thanks for coming on theCUBE, thanks for-- >> Thanks for having us. >> For making time to come over. This is theCUBE, breaking out the start-up action sharing the hot start-ups here, that really are a good position in the marketplace, as the generation of the infrastructure changes. It's a whole new ballgame. Global development platform, called the Internet. The new Internet. It's decentralized, we even get into Blockchain, we want to try that a little later, maybe another segment. It's theCUBE in New York City. More after this short break.

>> Narrator: Live from midtown Manhattan, it's theCUBE. Covering BigData, New York City, 2017. Brought to you by SiliconANGLE Media and its ecosystem sponsors. >> Welcome back everyone, we're here live in New York City for theCUBE's coverage of BigData NYC, our event we've been running for five years, been covering BigData space for eight years, since 2010 when it was Hadoop World, Strata Conference, Strata Hadoop, Strata Data, soon to be called Strata AI, just a few. We've been theCUBE for all eight years. Here, live in New York, I'm John Furrier. Our next guest is Murthy Mathiprakasam, who is the Director of Product Marketing at Informatica. Cube alumni has been on many times, we cover Informatica World, every year. Great to see you, thanks for coming by and coming in. >> Great to see you. >> You guys do data, so there's not a lot of recycling going on in the data because we've been talking about it all week, total transformation, but the undercurrent has been a lot of AI, AI this, and you guys have the CLAIRE product, doing a lot of things there. But outside of the AI, the undertone is cloud, cloud, cloud. Governance, governance, governance. There's two kind of the drivers I'm seeing as the force of this week is, a lot of people trying to get their act together on those two fronts and you can kind of see the scabs on the industry, people, some people haven't been paying attention. And they're weak in the area. Cloud is absolutely going to be driving the BigData world, 'cause data is horizontal. Cloud's the power source that you guys have been on that. What's your thoughts, what other drivers encourage you? (mumbles) what I'm saying and what else did I miss? Security is obviously in there, but-- >> Absolutely, no, so I think you're exactly right on. So obviously governments security is a big deal. Largely being driven by the GDPR regulation, it's happening in Europe. But, I mean every company today is global, so. Everybody's essentially affected by it. So, I think data until now has always been a kind of opportunistic thing, that there's a couple guys and their organizations were looking at it as oh, let's do some experimentation. Let's do something interesting here. Now, it's becoming government managed so I think there's a lot of organizations who are, like, to your point, getting their act together, and that's driving a lot of demand for data management projects. So now, people say, well, if I got to get my act together, I don't have to hire armies of people to do it, let me look for automated machine learning based ways of doing it. So that they can actually deliver on their audit reports that they need to deliver on, and ensure the compliance that they need to ensure, but do it in a very scalable way. >> I've been kind of joking all week, and I kind of had this meme in my head, so I've been pounding on it all week, calling it the tool shed problem. The tool shed problem is, everyone's got these tools. They throw them into the tool shed. They bought a hammer and the company that sells them the hammer is trying to turn it to a lawnmower, right? You can't mow your lawn with a hammer, it's not going to work, and so this, these tools are great but it defines work. What you do, but, the platforming issue is a huge one. And you start to see people who took that view. You guys were one of them because in a platform centric view with tools that are enabled, to be highly productive. You don't have to worry about new things like a government's policy, the GDPR that might pop up, or the next Equifax that's around the corner. There's probably two or three of them going on right now. So, that's an impact, the data, who uses it, how it's used, and who's at fault or whatever. So, how does a company deal with that? And machine learning has proven to be a great horse that a lot of people are riding right now. You guys are doing it, how does a customer deal with that tsunami of potential threats? Architecture challenges, what is your solution, how do you talk about that? >> Well, I think machine learning, you know, up until now has been seen as the kind of, nice to have, and I think that very quickly, it's going to become a must have. Because, exactly like you're saying, it really is a tsunami. I mean, you could see people who are nervous about the fact that I mean, there's different estimates. It's like 40% growth in data assets from most organizations every year. So, you can try to get around this somehow with one of these (mumbles) tools or something. But at some point, something is going to break, either you just don't, run out of manpower, you can't train the manpower, people start leaving. whatever the operational challenges are, it just isn't going to scale. Machine learning is the only approach. It is absolutely the only approach that actually ensures that you can maintain data for these kind of defensive reasons like you're saying. The structure and compliance, but also the kind of offensive opportunistic reasons, and do it scalably, 'cause there's just no other way mathematically speaking, that when the data is growing 40% a year, just throwing a bunch of tools at it just doesn't work. >> Yeah, I would just amplify and look right in the camera, say, if you're not on machine learning, you're out of business. That's a straight up obvious trend, 'cause that's a precursor to AI, real AI. Alright, let's get down to data management, so when people throw around data management, it's like, oh yeah we've got some data management. There are challenges with that. You guys have been there from day one. But now if you take it out in the future, how do you guys provide the data management in a totally cloud world where now the customer certainly has public and private, or on premise but theirs might have multi cloud? So now, comes a land grab for the data layer, how do you guys play in that? >> Well, I think it's a great opportunity for these kind of middle work platforms that actually do span multiple clouds, that can span the internal environments. So, I'll give you an example. Yesterday we actually had a customer speaking at Astrada here, and he was talking about from him, the cloud is really just a natural extension of what they're already doing, because they already have a sophisticated data practice. This is a large financial services organization, and he's saying well now the data isn't all inside, some of it's outside, you've got partners, who've got data outside. How do we get to that data? Clearly, the cloud is the path for doing that. So, the fact that the cloud is a national extension a lot of organizations were already doing internally means they don't want to have a completely different approach to the data management. They want to have a consistent, simple, systematic repeatable approach to the data management that spans, as you said, on premise in the cloud. That's why I think the opportunity of a very mature and sophisticated platform because you're not rewriting and re-platforming for every new, is it AWS, is it Azure? Is it something on premise? You just want something that works, that shields you from the underlying infrastructure. >> So I put my skeptic hat on for a second and challenge you on this, because this I think is fundamental. Whether it's real or not, it's perceived, maybe in the back of the mind of the CXO or the CDO, whoever is enabled to make these big calls. If they have the keys to the kingdom in Informatica, I'm going to get locked in. So, this is a deep fear. People wake up with nightmares in the enterprise, they've seen locked in before. How do you explain that to a customer that you're going to be an enabling opportunity for them, not a lock in and foreclosing future benefits. Especially if I have an unknown scenario called multi-cloud. I mean, no one's really doing multi-cloud let's face it. I mean, I have multiple clouds with stuff on it, >> At least not intentionally. Sometimes you got a line of businesses and doing things, but absolutely I get it. >> No one's really moving workloads dynamically between clouds in real time. Maybe a few people doing some hacks, but for the most part of course, not a standard practice. >> Right. >> But they want it to be. >> Absolutely. >> So that's the future. From today, how do you preserve that position with the customer where you say hey we're going to add value, but we're not going to lock you in? >> So the whole premise again of, I mean, this goes back to classic three tier models of how you think about technology stacks, right? There's an infrastructure layer, there's a platform layer, there's an analytics layer and the whole premise of the middle of the layer, the platform layer, is that it enables flexibility in the other two layers. It's precisely when you don't have something that's kind of intermediating the data and the use of the data, that's when you run into challenges with flexibility and with data being locked in the data store. But you're absolutely right. We had dinner with a bunch of our customers last night. They were talking about they'd essentially evaluated every version of sort of BigData platform and data infrastructure platform right? And why? It was because they were a large organization and your different teams start stuff and they had to compute them out and stuff. And I was like that must have been pretty hard for you guys. Now what we were using Informatica, so it didn't really matter where the data was, we were still doing everything as far as the data management goes from a consistent layer and we integrate with all those different platforms. >> John: So you didn't get in the way? >> We didn't get in the way. >> You've actually facilitated. >> We are facilitating increased flexibility. Because without a layer like that, a fabric, or whatever you want to call it a data platform that's facilitating this the complexity's going to get very, very crazy very soon. If it hasn't already. The number of infrastructure platforms that are available like you said, on premise and on the cloud now, keeps growing. The number of analytical tools that are available is also growing. And all this is amazing innovation by the way. This is all great stuff, but to your point about it if your the chief officer of an organization going, I got to get this thing figured out somehow. I need some sanity, that's really the purpose of-- >> They just don't want to know the tool for tool's sake, they need to have it be purposeful. >> And that's why this machine learning aspect is very, very critical because I was thinking about an analogy just like you were and I was thinking, in a way you can think of data managing as sort of cleaning stuff up and there are people that have brooms and mops and all these different tools. Well, we are bringing a Roomba to market, right? Because you don't want to just create tools that transfer the laborer around, which is a little bit of what's going on. You want to actually get the laborer out of the equation, so that the people are focused on the context, business strategy and the data management is sort of cleaning itself. It's doing the work for you. That's really what Informatica's vision is. It's about being a kind of enterprise cloud data management vendor that is leveraging AI under the hood so that you can sort of set it and forget it. A lot of this ingestion and the cleansing, telling annals what data they should be looking for. All the stuff is just happening in an automated way and you're not in this total chaos. >> And that can be some tools will be sitting in the back for a long time. In my tool shed, when I had one back in a big enough property back east. No one has tool sheds by the way. No one does any gardening. The issue is in the day, I need to have a reliable partner. So I want you to take a minute and explain to the folks who aren't yet Informatica customers why they should be and the Informatica customers why they should stay with Informatica. >> Absolutely, so certainly the ones we have, a very loyal customer base. In fact the guy who was presenting with us yesterday, he said he's been with Informatica since 1999, going through various versions of our products and adopting new innovations. So we have a very loyal customer base, so I think that loyalty itself speaks for itself as well. As far as net new customers, I think that in a world of this increasing data complexity, it's exactly what you were saying, you need to find an approach that is going to scale. I keep hearing this word from the chief data officer, I kind of got something some going on today, I don't know how I scale it. How is this going to work in 2018 and 2019, in 2025? And it's just daunting for some of these guys. Especially going back to your point about compliance, right? So it's one thing if you have data sitting around, data so to speak, that you're not using it. But god forbid now, you got legal and regulatory concerns around it as well. So you have to get your arms around the data and that's precisely where Informatica can help because we've actually thought through these problems and we've talked about them. >> Most of them were a problem you solved because at the end of the day, we were talking about problems that have massive importance, big time consequences people can actually quantify. >> That's right. >> So what specific problem highest level do you solve is the most important, has the most consequences? >> Everything from ingestion of raw data sets from wherever like you said, in the cloud on premise, all the way through all the processes you need to make it fully usable. And we view that as one problem. There's other vendors who think that one aspect of that is a problem and it is worth solving. We really think, look at the end of the day, you got raw stuff and you have to turn it into useful stuff. Everything in there has to happen, so we might as well just give you everything and be very, very good at doing all those things. And so that's what we call enterprise cloud data management. It's everything from raw material to finished goods of insights. We want to be able to provide that in a consistent integrated and machine learning integrate it. >> Well you guys have a loyal customer base but to be fair and you kind of have to acknowledge that there is a point in time and not throw Informatica's away the big customers, big engagements. But there was a time in Informatica's history where you went private. There was some new management came in. There was a moment where the boat was taking on water, right? And you could almost look at it and say, hmm, you know, we're in this space. You guys retooled around that. Success to the team. Took it to another dimension. So that's the key thing. You know a lot of the companies become big and it's hard to get rid of. So the question is that's a statement. I think you guys done a great job. Yet, the boat might have taken on water, that's my opinion, but you can probably debate that. But I think as you get mature and you're in public, you just went private. But here's the thing, you guys have had a good product chop in Informatica, so I got to ask you the question. What cool things are you doing? Because remember, cool shiny new toys help put a little flash and glam on the nuts and bolts that scales. What are you guys doing? I know you just announced claire, some AI stuff. What's the hot stuff you're doing that's adding value? >> Yeah, absolutely, first of all, this kind of addresses your water comment as well. So we are probably one of the few vendors that spends almost about $200 million in R and D. And that hasn't changed through the acquisition. If anything, I think it actually increased a little bit because now our investors are even more committed to innovation. >> Well you're more nimble in private. A lot more nimble. >> Absolutely, a lot more ideas that are coming to the forefront. So there's never been any water just to be clear. But to answer your follow on question about some examples of this innovation. So I think Ahmed yesterday talked about some of our recent release as well but we really just keep pushing on this idea of, I know I keep saying this but it's this whole machine learning approach here of how can we learn more about the data? So one of the features, I'll give you an example, is if we can actually go look at a file and if we spot like a name and an address and some order information, that probably is a customer, right? And we know that right, because we've seen past data sets. So, there's examples of this pattern matching where you don't even have to have data that's filled out. And this is increasingly the way the data looks we are not dealing with relational tables anymore it's JSON files, it's web blogs, XML files, all of that data that you had to have that data scientists go through and parse and sift through, we just automatically recognize it now. If we can look for the data and understand it, we can match it. >> Put that in context in the order of benefits that, from the old way versus the current way, what's the pain levels? One versus the other, can you put context around that? In terms of, it's pretty significant. >> It's huge because again, back to this sort of volume and variety of data that people are trying to get into systems and do it very rapidly. I'll give you a really tangible customer case. So, this is a customer that presented at Informatica World a couple months ago. It's Jewelry TV, I can actually tell you the name. So there are one of these online kind of shopping sites and they've got a TV program that goes with the online site. So what they do is obviously when you promote something on TV, your orders go up online, right? They wanted to flip it around and they said, look, let's look at the web logs of the traffic that's on the website and then go promote that on the TV program. Because then you get a closed loop and start to have this explosion of sales. So they used Informatica, didn't have to do any of this hand coding. They just build this very quickly and with the graphical user interface that we provide, it leverages sparks streaming under the hood. So they are using all these technologies under the hood, they just didn't have to do any of the manual coding. Got this thing out in a couple days and it works. And they have been able to measure it and they're actually driving increased sales by taking the data and just getting it out to the people that need to see the data very, very quickly. So that's an example of a use case where this isn't just to your point about is this a small, incremental type of thing. No, there is a lot of money behind data if you can actually put it to good use. >> The consequences are grave and I think you've seen more and more, I mean the hacks just amplify it over and over again. It's not a cost center when you think about it. It has to be somehow configured differently as a profit center, even though it might not drive top line revenue directly like an app or anything else. It's not a cost center. If anything it will be treated as a profit center because you get hacked or someone's data is misused, you can be out of business. There is no profit. Look at the results of these hacks. >> The defensive argument is going to become very, very strong as these regulations come out. But, let's be clear, we work with a lot of the most advanced customers. There are people making money off of this. It can be a top line driver-- >> No it should be, it should be. That's exactly the mindset. So the final question for you before we break. I know we're out of time here. There are some chief data officers that are enabled, some aren't and that's just my observation. I don't want to pidgeonhole anyone, but some are enable to really drive change, some are just figureheads that are just managing the compliance risk and work for the CFO and say no to everything. I'm over-generalizing. But that's essentially how I see it. What's the problem with that? Because the cost center issue has, we've seen this moving before in the security business. Security should not be part of IT. That's it's own deal. >> Exactly. >> So we're kind of, this is kind of smoke, but we're coming out of the jungle here. Your thoughts on that. >> Yeah, you're absolutely right. We see a variety of models. We can see the evolution of those models and it's also very contextual to different industries. There are industries that are inherently more regulated, so that's why you're seeing the data people maybe more in those cost center areas that are focused on regulations and things like that. There's other industries that are a lot more consumer oriented. So for them, it makes more sense to have the data people be in a department that seems more revenue basing. So it's not entirely random. There are some reasons, that's not to say that's not the right model moving forward, but someday, you never know. There is a reason why this role became a CXO in the first place. Maybe it is somebody who reports to the CEO and they really view the data department as a strategic function. And it might take a while to get there, but I don't think it's going to take a long time. Again, we're talking about 40% growth in the data and these guys are realizing that now and I think we're going to see very quickly people moving out of the whole tool shed model, and moving to very systematic, repeatable practices. Sophisticated middleware platforms and-- >> As we say don't be a tool, be a platform. Murphy thanks so much for coming on to theCUBE, we really appreciate it. What's going on in Informatica real quick. Things good? >> Things are great. >> Good, awesome. Live from New York, this is theCUBE here at BigData NYC more live coverage continuing day three after this short break. (digital music)

>> Live from Midtown Manhattan, it's the CUBE. Covering Big Data New York City 2017. Brought to you by SiliconANGLE Media and its ecosystem sponsors. >> Fifth year of coverage of our own event Big Data NYC where we cover all the action in New York City. For this week in big data, in conjunction with Strata Data which was originally Hadoop World in 2010. We've been covering it for eight years. It became Strata Conference, Strata Hadoop, now called Strata Data. Will probably called Strata AI tomorrow. Who knows, but certainly the trends are going in that direction. I'm John Furrier, your co-host. Our next guest here in New York City is Sergei Rabotai, who is the Head of Business Development at InData Labs from Belarus. In town, doing some biz dev in the big data ecosystem. Welcome to theCUBE. >> Yeah. Good morning. >> Great to have you. So, obviously Belarus is becoming known as the Silicon Valley of Eastern Europe. A lot of great talent. We're seeing that really explode. A lot of great stuff going on globally, even though there's a lot of stuff, you know GDPR and all these other things happening. It's clearly a global economy with tech. Silicon Valley still is magical. I live there in Palo Alto but you're starting to see peering points within these ecosystems of entrepreneurship and now big companies are taking advantage of it as well. What do you guys do? I mean you're in the middle of that. What is InData Labs do in context of all this? >> Well, InData Labs is a full stack data science company. Which means that we provide professional services for data strategy, big data engineering and the data science. So, yeah, like you just said, we are based - my team is based in Minsk, Belarus. We are about 40 people strong at the moment. And in our recent years we have been very successful starting this business and we have been getting customers from all over the world, including United States, Great Britain, and European Union. The company was launched about four years ago and very important thing, that it was launched by two tech leaders who come from very data-driven industries. Our CEO, Ilya Kirillov, has been running several EdTech companies for many years. Our second founder, Marat Karpeko, has been holding C-Level positions in one of the most successful gaming companies in the world. >> John: So they know data. They're data guys. >> Yeah they're data guys. They know data from different aspects and that brings synergy to our business. >> You guys bring that expertise now into professional services for us. Give me an example of some of the things someone might want to call you up on, because the thing we're hearing here in New York City this week is look, we need more data sciences and they got to be more productive. They're spending way too much time wrangling and doing stuff that they shouldn't be doing. In the old days, sysadmins were built to let people be productive and they ran the infrastructure. That's not what data scientists should be doing. They're the users. There's a level of setting things up and then there's a level of provisioning, it's actually data assets, but then the data scientists just want to do their job. How do you help companies do that? >> Well I would probably, if I take all of our activities, I would split them into two big parts. First of all, we are helping big companies, who already have a lot of data. We help them in managing this data more effectively. We help them with predictive analytics. We help them with, helping them build the churn prediction and user segmentation solutions. We have been recently involved into several natural language processing projects. In one of our successful key studies we helped one of the largest gaming companies to automate their customer feedback processing. So, like, a couple years ago they were working manually with their customer feedback and we built them a tool that allows them to instantly get the sentiment of what the user says. It's kind of like a voice of a customer, which means they can be more effective in developing new things for their games. So, we-- >> So what would someone engage? I'm just trying to peg a order of magnitude of the levels of engagements you do. Startups come in? Is it big companies? What kind of size scoped work do you do? >> So I would say at the moment we work with startups, but it's a bit of a different approach than we have with big or well-established companies. When startups typically approach us with asking to help them implement some brand new technologies like neural networks or deep learning. So they want to be effective from the start. They want to use the cutting edge technology to be more attractive, to provide a better value on the market and just to be effective and to be a successful business from the start. The other part, the well-established companies, who already have the data but they understand that so far their data might not be used that effectively as it should have been used. Therefore, they approach us with a request to help them to get more insights out of the data. Let's say, implement some machine learning that can help them. >> How about larger companies? What kind of projects do you work for them? >> It could be a typical project like churn prediction, that is very actual for the companies who have got a lot of customer data. Then it could be companies from such industries like betting industry, where churn is a very big issue. And, the same probably applies to companies who do trading. >> So is scale one of the things you differentiate around? It sounds like your founders have an EdTech background obviously must be a larger, large data set. Is your profile of engagements large scale? Is it ... I'm just trying to get a handle of if someone's watching who, what is the kind of engagements people should be calling you for? Give us an example of that. >> Like, let's say there is a company who has got a lot of customer data, has got some products and they have a problem of churn, or they have a problem of segmenting their customers so they can later address the specific segments of the customers with the right offers at the right time and through the right marketing channel. Then it could be customers or requests where natural text processing is required where we have to automate some understanding of the written or spoken text. Then I should say that we have been getting recently some requests where computer vision skills are required. I think the first stage of AI being really intelligent was the speech recognition and I think nowadays we manage to reach to the level of what we earlier saw in fantastic movies or sci-fi movies. Computer vision is going to be the next leap in all that AI buzz we're having at the moment. >> So you solve, the problem that you solve for customers is data problems. If they're swimming in a lot of data, you can help them. >> Sergei: Yep. >> If they actually want to make that data do things that are cutting edge, you guys can help them. >> Sergei: Yeah. That's-- >> Alright, so here's a question for you. I mean, Belarus has obviously got good things going on. I've heard the press that you guys have been getting, the whole area, and you guys in particular. So I'm a buyer, one of the questions I might ask is "Hey Sergei, how do I know that you'll keep that talent because the churn is always a big problem. I've dealt with outsourcing before and in the US it's hard to keep talent but I've heard there's a churn." How do you guys keep the talent in the country? How do you keep talent on the projects? Is there certain economic rules over there? What's happening in Belarus? Give us the economical. >> Yeah, so, basically what you're saying. The churn problem has always been known for companies who have their development teams in Asian regions. That's a known problem because I have a lot of meetings with clients in the UK and the US, potential prospects, I would say. So they say it is a problem for them. With Belarus, I don't think we have that because from what I know, we have an average churn of under 10 percent. That's the figures across the industry. In smaller companies, the churn is even less and there are specific reasons for that. First of all, that due to Belarusian mentality, we always try to keep to a job that we're having. Yeah? So we do not-- >> John: That's a cultural thing. >> That's just the cultural thing. We do not ... >> You honor, you honor a code, if you will. >> Yeah. >> Okay. >> So, that's one of the things. Another thing is that Belarusian IT industry is very small. We have, I would say, no more than 40 thousand people being involved in different IT companies. The community is very small, so if somebody is hopping jobs from one job to another, it is going to be known and this person is not likely to have like, a good career. >> So job hoppers is kind of like a code of community, honor. Silicon Valley works that way too, by the way. >> Yeah. >> You get identified, that's who you are. >> Yeah. And so nowadays-- >> Economic tax breaks going on over there? What's the government to get involved? >> One of the key things is, the special tax and legal regulations that Belarus has got at the moment. I can definitely say that there is no country in the world that has got the same tax preferences, and the same support from the government. If a Belarusian company, IT company, becomes a part of Belarusian High Tech Park it means the company becomes automatically exempt from BET tax, corporate income tax. The employees of that company having the reliefs on their income, personal income tax rate, and there are a lot more reliefs that make the talent stay in the country. Having this relief for the IT business allows the companies to provide better working conditions for the employees and stop the people from migrating to other parts of the world. That's what we have. >> Sort of created an environment where there's not a lot of migration out of the area. The tech community kind of does it's own policing of behavior for innovation. >> Yeah but I think before those initiatives were adopted there was a certain percentage of people migrating but I think that nowadays even if it happens, yes, you're right, it's not that substantial. >> Great. Tell us ... Great overview of the company and congratulations, it's a good opportunity for folks watching to explore new areas of talent, especially ones that have the work ethic and knowledge you guys have over there. New York here, there's codes here too. Get the job done. Be on time. What's your experience like in New York here? What's your goal this week? What's some of the meetings you're having? Share with the folks kind of your game plan for Big Data NYC. >> Well, yeah, I've really enjoyed my stay here. It, so far, has been a very enjoyable experience. From the business perspective, I had over 10 meetings with the prospective customers. And we are likely to have follow-ups coming in the next couple of weeks. I can definitely say there is a great demand for professional services. You can see that if you go to whichever center you can see there's a lot of jobs being posted on the job boards. It means that there is lack of knowledge here in the US, yeah? One more important thing that I wanted to share with you from my personal observations that USA, UK and maybe Nordic countries, they have very, very strong background for creating the business ideas but Eastern Europe or Eastern European countries and Belarus in particular, they are very strong in actually implementing those ideas. >> Building them. >> Yes, building them. I think we have lots of synergies and we can ... we can ... >> John: Great. >> We can work together. I also got some meetings with our existing customers here in the US and so far we had good experiences. I can see that New York is moving fast. I travel a lot. I've been to over 40 countries in the previous five years and I just ... New York is different. >> It's fun. >> Different. Even different from many other cities in the US. >> Lot of banks are here. Lot of business in New York. New York is a great town. Love New York City. It's one of my favorites. Love coming here as I grew up right across the river in New Jersey. >> Yeah. But, great town, obviously California, Palo Alto, >> Yeah. >> Is a little more softer in terms of weather, but they have a culture there too. Sounds a lot like what's going on in Belarus, so congratulations. If we get some business for you, should we give them theCUBE discount, tell them John sent you and you get 10 percent off? Alright? >> Alright, yes. Sounds great. We can make it a good deal. (laughter) >> Tell them John sent you, you get 10% off. No I'm only kidding because it's services. Congratulations. Final question. What's the number one thing that people are buying for service from you guys? Number one thing. What's the most requested service you provide? >> The most requested services ... First of all, many customers they understand that they have got a lot of data. They want to do something with their data. But before you actually do some implementation you have to do a lot of discovery or preparatory work. I would say, no matter how we end up with a customer, this stage is basically ... The idea of that stage is to identify the ways data science can be implemented and can provide benefits to the business. That's the most important. I think that, like, 95 percent of the customers they approach us with this thing in the first place. And based on the results of that preparatory stage we can then advise the customers. What can they do? Or how they can actually benefit from the existing data? Or what other things they should collect in order to make their business more effective. >> Sergei, thanks for coming on. Belarus has got a lot of builders there. Check 'em out. >> Thanks a lot. >> Builders are critical in this new world. Lots of them with clout, a lot of great opportunities. A lot of builders in Belarus. This is theCUBE, bringing you all the action from New York City. More after this short break. We'll be right back. (theme music) (no audio) >> Hi, I'm John Furrier, the co-founder of SiliconANGLE Media and co-host of theCUBE. I've been in the tech ...

>> Coming back when we were a DOS spreadsheet company. I did a short stint at Microsoft and then joined Frank Quattrone when he spun out of Morgan Stanley to create what would become the number three tech investment (upbeat music) >> Host: Live from mid-town Manhattan, it's theCUBE covering the BigData New York City 2017. Brought to you by SiliconANGLE Media and its ecosystem sponsors. (upbeat electronic music) >> Welcome back, everyone. We're here, live, on day two of our three days of coverage of BigData NYC. This is our event that we put on every year. It's our fifth year doing BigData NYC in conjunction with Hadoop World which evolved into Strata Conference, which evolved into Strata Hadoop, now called Strata Data. Probably next year will be called Strata AI, but we're still theCUBE, we'll always be theCUBE and this our BigData NYC, our eighth year covering the BigData world since Hadoop World. And then as Hortonworks came on we started covering Hortonworks' data summit. >> Arun: DataWorks Summit. >> DataWorks Summit. Arun Murthy, my next guest, Co-Founder and Chief Product Officer of Hortonworks. Great to see you, looking good. >> Likewise, thank you. Thanks for having me. >> Boy, what a journey. Hadoop, years ago, >> 12 years now. >> I still remember, you guys came out of Yahoo, you guys put Hortonworks together and then since, gone public, first to go public, then Cloudera just went public. So, the Hadoop World is pretty much out there, everyone knows where it's at, it's got to nice use case, but the whole world's moved around it. You guys have been, really the first of the Hadoop players, before ever Cloudera, on this notion of data in flight, or, I call, real-time data but I think, you guys call it data-in-motion. Batch, we all know what Batch does, a lot of things to do with Batch, you can optimize it, it's not going anywhere, it's going to grow. Real-time data-in-motion's a huge deal. Give us the update. >> Absolutely, you know, we've obviously been in this space, personally, I've been in this for about 12 years now. So, we've had a lot of time to think about it. >> Host: Since you were 12? >> Yeah. (laughs) Almost. Probably look like it. So, back in 2014 and '15 when we, sort of, went public and we're started looking around, the thesis always was, yes, Hadoop is important, we're going to love you to manage lots and lots of data, but a lot of the stuff we've done since the beginning, starting with YARN and so on, was really enable the use cases beyond the whole traditional transactions and analytics. And Drop, our CO calls it, his vision's always been we've got to get into a pre-transactional world, if you will, rather than the post-transactional analytics and BIN and so on. So that's where it started. And increasingly, the obvious next step was to say, look enterprises want to be able to get insights from data, but they also want, increasingly, they want to get insights and they want to deal with it in real-time. You know while you're in you shopping cart. They want to make sure you don't abandon your shopping cart. If you were sitting at at retailer and you're on an island and you're about to walk away from a dress, you want to be able to do something about it. So, this notion of real-time is really important because it helps the enterprise connect with the customer at the point of action, if you will, and provide value right away rather than having to try to do this post-transaction. So, it's been a really important journey. We went and bought this company called Onyara, which is a bunch of geeks like us who started off with the government, built this batching NiFi thing, huge community. Its just, like, taking off at this point. It's been a fantastic thing to join hands and join the team and keep pushing in the whole streaming data style. >> There's a real, I don't mean to tangent but I do since you brought up community I wanted to bring this up. It's been the theme here this week. It's more and more obvious that the community role is becoming central, beyond open-source. We all know open-source, standing on the shoulders before us, you know. And Linux Foundation showing code numbers hitting up from $64 million to billions in the next five, ten years, exponential growth of new code coming in. So open-source certainly blew me. But now community is translating to things you start to see blockchain, very community based. That's a whole new currency market that's changing the financial landscape, ICOs and what-not, that's just one data point. Businesses, marketing communities, you're starting to see data as a fundamental thing around communities. And certainly it's going to change the vendor landscape. So you guys compare to, Cloudera and others have always been community driven. >> Yeah our philosophy has been simple. You know, more eyes and more hands are better than fewer. And it's been one of the cornerstones of our founding thesis, if you will. And you saw how that's gone on over course of six years we've been around. Super-excited to have someone like IBM join hands, it happened at DataWorks Summit in San Jose. That announcement, again, is a reflection of the fact that we've been very, very community driven and very, very ecosystem driven. >> Communities are fundamentally built on trust and partnering. >> Arun: Exactly >> Coding is pretty obvious, you code with your friends. You code with people who are good, they become your friends. There's an honor system among you. You're starting to see that in the corporate deals. So explain the dynamic there and some of the successes that you guys have had on the product side where one plus one equals more than two. One plus one equals five or three. >> You know IBM has been a great example. They've decided to focus on their strengths which is around Watson and machine learning and for us to focus on our strengths around data management, infrastructure, cloud and so on. So this combination of DSX, which is their data science work experience, along with Hortonworks is really powerful. We are seeing that over and over again. Just yesterday we announced the whole Dataplane thing, we were super excited about it. And now to get IBM to say, we'll get in our technologies and our IP, big data, whether it's big Quality or big Insights or big SEQUEL, and the word has been phenomenal. >> Well the Dataplane announcement, finally people who know me know that I hate the term data lake. I always said it's always been a data ocean. So I get redemption because now the data lakes, now it's admitting it's a horrible name but just saying stitching together the data lakes, Which is essentially a data ocean. Data lakes are out there and you can form these data lakes, or data sets, batch, whatever, but connecting them and integrating them is a huge issue, especially with security. >> And a lot of it is, it's also just pragmatism. We start off with this notion of data lake and say, hey, you got too many silos inside the enterprise in one data center, you want to put them together. But then increasingly, as Hadoop has become more and more mainstream, I can't remember the last time I had to explain what Hadoop is to somebody. As it has become mainstream, couple things have happened. One is, we talked about streaming data. We see all the time, especially with HTF. We have customers streaming data from autonomous cars. You have customers streaming from security cameras. You can put a small minify agent in a security camera or smart phone and can stream it all the way back. Then you get into physics. You're up against the laws of physics. If you have a security camera in Japan, why would you want to move it all the way to California and process it. You'd rather do it right there, right? So with this notion of a regional data center becomes really important. >> And that talks to the Edge as well. >> Exactly, right. So you want to have something in Japan that collects all of the security cameras in Tokyo, and you do analysis and push what you want back here, right. So that's physics. The other thing we are increasingly seeing is with data sovereignty rules especially things like GDPR, there's now regulation reasons where data has to naturally stay in different regions. Customer data from Germany cannot move to France or visa versa, right. >> Data governance is a huge issue and this is the problem I have with data governance. I am really looking for a solution so if you can illuminate this it would be great. So there is going to be an Equifax out there again. >> Arun: Oh, for sure. >> And the problem is, is that going to force some regulation change? So what we see is, certainly on the mugi bond side, I see it personally is that, you can almost see that something else will happen that'll force some policy regulation or governance. You don't want to screw up your data. You also don't want to rewrite your applications or rewrite you machine learning algorithms. So there's a lot of waste potential by not structuring the data properly. Can you comment on what's the preferred path? >> Absolutely, and that's why we've been working on things like Dataplane for almost a couple of years now. We is to say, you have to have data and policies which make sense, given a context. And the context is going to change by application, by usage, by compliance, by law. So, now to manage 20, 30, 50 a 100 data lakes, would it be better, not saying lakes, data ponds, >> [Host} Any Data. >> Any data >> Any data pool, stream, river, ocean, whatever. (laughs) >> Jacuzzis. Data jacuzzis, right. So what you want to do is want a holistic fabric, I like the term, you know Forrester uses, they call it the fabric. >> Host: Data fabric. >> Data fabric, right? You want a fabric over these so you can actually control and maintain governance and security centrally, but apply it with context. Last not least, is you want to do this whether it's on frame or on the cloud, or multi-cloud. So we've been working with a bank. They were probably based in Germany but for GDPR they had to stand up something in France now. They had French customers, but for a bunch of new reasons, regulation reasons, they had to sign up something in France. So they bring their own data center, then they had only the cloud provider, right, who I won't name. And they were great, things are working well. Now they want to expand the similar offering to customers in Asia. It turns out their favorite cloud vendor was not available in Asia or they were not available in time frame which made sense for the offering. So they had to go with cloud vendor two. So now although each of the vendors will do their job in terms of giving you all the security and governance and so on, the fact that you are to manage it three ways, one for OnFrame, one for cloud vendor A and B, was really hard, too hard for them. So this notion of a fabric across these things, which is Dataplane. And that, by the way, is based by all the open source technologies we love like Atlas and Ranger. By the way, that is also what IBM is betting on and what the entire ecosystem, but it seems like a no-brainer at this point. That was the kind of reason why we foresaw the need for something like a Dataplane and obviously couldn't be more excited to have something like that in the market today as a net new service that people can use. >> You get the catalogs, security controls, data integration. >> Arun: Exactly. >> Then you get the cloud, whatever, pick your cloud scenario, you can do that. Killer architecture, I liked it a lot. I guess the question I have for you personally is what's driving the product decisions at Hortonworks? And the second part of that question is, how does that change your ecosystem engagement? Because you guys have been very friendly in a partnering sense and also very good with the ecosystem. How are you guys deciding the product strategies? Does it bubble up from the community? Is there an ivory tower, let's go take that hill? >> It's both, because what typically happens is obviously we've been in the community now for a long time. Working publicly now with well over 1,000 customers not only puts a lot of responsibility on our shoulders but it's also very nice because it gives us a vantage point which is unique. That's number one. The second one we see is being in the community, also we see the fact that people are starting to solve the problems. So it's another elementary for us. So you have one as the enterprise side, we see what the enterprises are facing which is kind of where Dataplane came in, but we also saw in the community where people are starting to ask us about hey, can you do multi-cluster Atlas? Or multi-cluster Ranger? Put two and two together and say there is a real need. >> So you get some consensus. >> You get some consensus, and you also see that on the enterprise side. Last not least is when went to friends like IBM and say hey we're doing this. This is where we can position this, right. So we can actually bring in IGSC, you can bring big Quality and bring all these type, >> [Host} So things had clicked with IBM? >> Exactly. >> Rob Thomas was thinking the same thing. Bring in the power system and the horsepower. >> Exactly, yep. We announced something, for example, we have been working with the power guys and NVIDIA, for deep learning, right. That sort of stuff is what clicks if you're in the community long enough, if you have the vantage point of the enterprise long enough, it feels like the two of them click. And that's frankly, my job. >> Great, and you've got obviously the landscape. The waves are coming in. So I've got to ask you, the big waves are coming in and you're seeing people starting to get hip with the couple of key things that they got to get their hands on. They need to have the big surfboards, metaphorically speaking. They got to have some good products, big emphasis on real value. Don't give me any hype, don't give me a head fake. You know, I buy, okay, AI Wash, and people can see right through that. Alright, that's clear. But AI's great. We all cheer for AI but the reality is, everyone knows that's pretty much b.s. except for core machine learning is on the front edge of innovation. So that's cool, but value. [Laughs] Hey I've got the integrate and operationalize my data so that's the big wave that's coming. Comment on the community piece because enterprises now are realizing as open source becomes the dominant source of value for them, they are now really going to the next level. It used to be like the emerging enterprises that knew open source. The guys will volunteer and they may not go deeper in the community. But now more people in the enterprises are in open source communities, they are recruiting from open source communities, and that's impacting their business. What's your advice for someone who's been in the community of open source? Lessons you've learned, what is the best practice, from your standpoint on philosophy, how to build into the community, how to build a community model. >> Yeah, I mean, the end of the day, my best advice is to say look, the community is defined by the people who contribute. So, you get advice if you contribute. Which means, if that's the fundamental truth. Which means you have to get your legal policies and so on to a point that you can actually start to let your employees contribute. That kicks off a flywheel, where you can actually go then recruit the best talent, because the best talent wants to stand out. Github is a resume now. It is not a word doc. If you don't allow them to build that resume they're not going to come by and it's just a fundamental truth. >> It's self governing, it's reality. >> It's reality, exactly. Right and we see that over and over again. It's taken time but it as with things, the flywheel has changed enough. >> A whole new generation's coming online. If you look at the young kids coming in now, it is an amazing environment. You've got TensorFlow, all this cool stuff happening. It's just amazing. >> You, know 20 years ago that wouldn't happen because the Googles of the world won't open source it. Now increasingly, >> The secret's out, open source works. >> Yeah, (laughs) shh. >> Tell everybody. You know they know already but, This is changing some of the how H.R. works and how people collaborate, >> And the policies around it. The legal policies around contribution so, >> Arun, great to see you. Congratulations. It's been fun to watch the Hortonworks journey. I want to appreciate you and Rob Bearden for supporting theCUBE here in BigData NYC. If is wasn't for Hortonworks and Rob Bearden and your support, theCUBE would not be part of the Strata Data, which we are not allowed to broadcast into, for the record. O'Reilly Media does not allow TheCube or our analysts inside their venue. They've excluded us and that's a bummer for them. They're a closed organization. But I want to thank Hortonworks and you guys for supporting us. >> Arun: Likewise. >> We really appreciate it. >> Arun: Thanks for having me back. >> Thanks and shout out to Rob Bearden. Good luck and CPO, it's a fun job, you know, not the pressure. I got a lot of pressure. A whole lot. >> Arun: Alright, thanks. >> More Cube coverage after this short break. (upbeat electronic music)