 Live from New York, it's theCUBE. Covering Big Data New York City 2016. Brought to you by headline sponsors, Cisco, IBM, NVIDIA, and our ecosystem sponsors. Now, here are your hosts, Dave Vellante and Peter Burris. New York City everybody, this is theCUBE, the worldwide leader in live tech coverage. And we're here at Big Data NYC, hashtag Big Data NYC, and we run Big Data NYC concurrent with Strata and Hadoop. What used to be Hadoop World and combined to be Strata Hadoop just down the street at the Javits Center. This is the seventh year of theCUBE covering what was Hadoop World, it was one of our first shows in 2010. And as I say, it's evolved into Strata. Interestingly, Hadoop is no longer the shiny new toy. We're going to talk about that. We have a panel of our experts here. Let me introduce them now. Peter Burris is the chief research officer of SiliconANGLE Media and has been leading up a lot of the work that we're doing in data, capital and data value. Jeff Frick is here, he's the general manager of theCUBE. Jeff has been running, it was just an awesome program this week and throughout the year. And of course we're joined by George Gilbert, who's a lead big data analyst at Wikibon. Gentlemen, good to see you again. Good day. Good night last night, Jeff. We had NVIDIA, was our co-host of an event, the party, Peter Burris was the host of that party. And we were talking about deep learning and AI and give us a little update on what that event was about. You know, it's just interesting how this continues to evolve. We were just talking about what is the hashtag? It's still Strata Hadoop. You know, the Hadoop is kind of going down in terms of importance, but there's this whole new thing with machine learning and artificial intelligence and driven by autonomous cars and all these new things. So again, this is new wave of innovation, new wave of opportunity. A ton of the people on the panel last night had left, you know, some of the leaders in the Hadoop space, if you will, to start new companies around new opportunities in real time and advanced analytics. So it's great. It's the Silicon Valley way, right? Onto the next big thing. Of course George, you also hosted a panel last night and we had a packed house. There were probably 150, 160 people here. They stayed, you know, through the content. In fact, most of them, many of them left after the content presentation and the DJ started going and the beer started, maybe they just wanted to see the debate and then plenty of people stayed. But so George, what are the big themes that we should be looking for here at Strata Hadoop this year? Well, there's some that sort of continue on from the last couple of shows that we've done where Hadoop was our shining new toy for a couple of years but it turned out that, you know, as Peter has articulated really well, there's a skill set involved in putting all the pieces together. We sort of broke apart what it used to take an integrated product like a DBMS to do all these advanced analytics and we took out the pieces so you could do a mix and match assembly and get essentially a pipeline of analytics that was just right for you and the unintended side effect was sort of a lot of complexity. That worked for the Googles, the LinkedIn's, the Twitters, the Ubers, the Airbnb's. The rest of the mainstream, they're still struggling to get their data lakes to be productive and the value of the data lake really was lower cost of ownership like per terabyte than a data warehouse. I will say, I think the shiny new toy that we're going to hear about a lot in this show is machine learning. But again, there's a huge amount of challenges to make that mature. So Peter, the return on investment in this space has been what Abhi met a joke is reduction in investment and how they've lowered the denominator as a way to get ROI, you know, okay, I'm going to put everything in the data lake and it's going to be cheaper to store it. Is there hope for this space in terms of extracting more value? Is there a data value play that is on the horizon, perhaps with machine learning and some of these other technologies that we're talking about or is it more of a business process and human capital issue? Well, so you're asking a couple of questions there, Dave. The first question you're asking in many respects is do these technologies make it the whole concept of data value clearer? And I think that's the first thing is that what companies are really trying to do with big data technologies is liberate the value of data, figure ways of translating data into value in new ways. Often because it allows them to understand the characteristics of the problems, a new. What the big data community is challenged to do, however, is to stop just chasing the new tool and instead start focusing on how these tools come together to actually solve real problems. So I think we're going to, to what George has said, we're going to see two things happen. We're going to see people talking about the new tool because they are trained and that's their habit is to keep chasing new tools. And we're going to talk about other people who are saying we've seen failures, we've seen lack of adoption, we've seen abandonment, but here's new ways of thinking about how these things come together in simpler, more straightforward combinations to actually solve these complex problems. The real object, look, this stuff is going to be some of the most important stuff that's ever come out of the computing industry. The whole concept of big data is incredibly important and it's going to have long lasting effects on not just this industry, but all industries. That's pretty obvious. The process of getting there is turning out to be a hell of a lot more painful than I think anybody thought it would be when we first started. And that has to become a theme of the show as well. How can we start doing this in a more simpler, disciplined way that more likely leads to predictable value? And part of that challenge has been the complexity within the ecosystem. You guys have written about that a lot. Last year, just exactly a year ago, we talked about there were cracks in that Hadoop armor. Said that nose of the plane is pointing up, but a lot of companies are losing altitude and when the funding dries up, uh-oh, look out. So we started to see some of that happen. There's been some consolidation. Although talent just is going public, so there's maybe some bright spots. What's your take on what's happening in the market, George, in the ecosystem? Give us the summary there. Well, you know, for many years, sort of digression, but an analogy. For many years, Amazon was funded with, or I should say their valuation was extremely high with marginal productivity and that gave them a cost of capital that was essentially zero and that's what allowed them to build out these, you know, this huge network of distribution centers. We were in a similar situation with many big data companies. They could grow their infrastructure in terms of their employee base, their skills, their distribution channels where with private funding, they got very, very high valuations. That changed towards the end of last year in the beginning of this year. And so where the assumption was, hey, our standard bearers are the Hadoop distro vendors and then there's the satellite vendors like Italian and others that help sort of bring data in or do advanced analytics. That's not taken for granted anymore and we're seeing people who got stuck, mainstream companies who got stuck trying to get value out of these, I should say, either out of Hadoop projects or more broadly big data, they're looking to say, well, maybe the answer isn't just open source so that I have freedom, maybe the answer is simplicity, even if it needs a little more lock-in. And that's what Peter was referring to, hey, maybe we'll take something that sort of comes all pre-integrated and has some scanner interfaces, let's say like an Iguaz, which we'll hear about more later this week, which is basically exadata for big data. Well, wasn't that MapR's play with the original DOOP distro? Yes, they are an example. And how's it working? I mean, it's hard to tell what's happening with Cloudera, they got a boatload of money from Intel and they seem to be doing okay, but suspect they're burning a lot of cash. We know what the story is that Hortonworks, they had to do, they announced another raise in last December, worst time in the world, you could do that, but they had to do it. And MapR, I don't know, a private company, hard to tell, you've seen some M&A Pentaho got taken out by Hitachi. But then again, Datamere's getting funding. So it seems like there's rip currents in all directions. There's still a lot of VC capital sloshing around. We go through these different phases where sometimes there's excess of capital chasing, not as many really high quality deals. Sometimes you don't have enough skilled people for these companies to hire. Right now we're still in the excess capital sort of overhang, but I think customers are starting to vote with their pocketbooks and say, look, it doesn't have to be open source. The big companies were okay with that. So what about the customer angle here? I mean, companies are struggling still with the data strategy, right? Well, let's put it this way, Dave. This has been a marketplace, so this is my second Hadoop-oriented Cube gig. And so I haven't done the previous seven that you guys have done, but I'll venture an opinion that for most of those previous seven, the talk was the chasing of new and cool tools. This is an ecosystem that has been chasing tools and hoping that those tools would ultimately lead to finding outcomes. What users are starting to do now increasingly is focus on the outcomes. They're chasing the outcomes. That's what they want. And whatever set of tools is available, they're starting to use. So while the ecosystem continues to chase tools and machine learning is an example of that, that's the habitual behavior of this marketplace. Customers are now stepping back and saying, I have been chasing tools. It hasn't generated the returns and results that I want. I now have to focus my time and attention on the outcomes, the capabilities required to get to those outcomes. And I'm going to start working more closely with suppliers and vendors that are capable of complimenting my skills and my capabilities so that I can achieve the outcomes that I want. The marketplace, in my opinion, in our opinion, is about to go through something of a not consequential shift. It's going to be subtle. I think it's going to start happening at this show in the conversations we have at this table. They focus on outcomes where tools are or are not achieving those outcomes and a reduction in the focus on tools per se. Yeah, you're right. This is a good observation, even though you weren't here. It was kind of, what is it? How can I get it to work? How can I get value out of it? Oh, well, I guess I can reduce the expense of my data warehouse. Yeah, but it's still not living up to the promises of a 365 degree view of the customer. Oh, we'll bring in Spark and it'll make it real time. Okay, well, how do I get value out of that? And it's just like this never ending tail chasing of to find value. You know, the real innovation, I think that we might start, we might start witnessing at this show is it's not the product side, it's the go to market side. You know, we're pretty close with what IBM keeps evolving in terms of their go to market. It's not about sort of putting up a console with 27 different services. It's training 2000 global business consultants on how to build solutions out of their components. And then even in the bigger group, the industry solutions group where they've built semi custom apps, those are outcomes. And I suspect over the next, you know, six to 12 months, we're going to be seeing more and more of companies like that. And I think we'll start to see the azures and the Amazons of the world try and get a little closer. Well, and IBM's making money in big data because they are focused on the outcomes. And speaking of IBM, we have an event with IBM tonight. I guess this afternoon, we got a bunch of IBM action going on. That's right. Tonight we're co-hosting a party with IBM. I love this co-hosting a party. It's working. It's working. One party which is not enough, Dave. We had another one. So tonight, the second half of the day is really this IBM signature event where they've got a lot of speakers talking about, again, the solutions play. And then they've got a party across the street at the other Mercantile Linux on 36th. Make sure I get my streets right. That starts at seven p. So that will be there. They'll have a whole another series of talks and panels will be carrying it live on theCUBE. So keep an eye if you can't make it. That's the Lincoln & Englebot TV. You'll catch it all there. Of course, we'll have it all up on demand. Yeah, and we're here at 37 Pillars, which is on 37th Street. It's 517, I think. 517. 37th Street. So stop by West 37th, obviously. Just for John Furrier, it's a driver from Javits. For me, it's three or four drivers. But stop by and see us. You're watching theCUBE. We're live from New York City. We'll be right back. Why wait for the-