 At Big Data SV 2014 is brought to you by headline sponsors, WAN Disco. We make Hadoop invincible and aptian, accelerating Big Data 2.0. Okay, welcome back everyone. This is SiliconANGLE and Wikibon's theCUBE, our flagship program. We go out to the events, extract the signal from the noise. I'm John Furrier, the founder of SiliconANGLE. I'm joined by Coast, Dave Vellante, co-founder of wikibon.org. And we're here in Silicon Valley, live in the heart of Big Data, Silicon Valley country, covering all the actions, startups, VCs, new big companies growing, and of course covering the Stratoconference right across the street. And my next guest is Monty's. We've been a CEO of Splice Machine. Welcome to theCUBE. Thank you very much. So you guys, hot startup, SQL on Hadoop, which has been a hot topic now going back two years. You guys just announced, just within the past 48 hours, $15 million of a Series B financing, which is validation. We were talking about that. Congratulations. And how do you feel? I feel great. Company is now on its second stage of financing, growing significantly. And we think we have a wonderful position where as everybody who is focused on SQL on Hadoop, they're so focused on analytics and having a repository for data science, whereas Splice Machine is really looking to power real-time applications with real transactions. And so we feel being the only real-time transactional SQL on Hadoop database, we've got a unique position in the marketplace to take advantage of that new financing. So we've covered, HADAP two years ago got fresh startup of the year. Dave, we've covered them. We remember we were at the Strata here and they won the startup competition. Some say the ballots were stuffed a little bit, but that's always the Twitter following. That's like getting your picture on that cover of Sports Illustrated, isn't it? It's like a kiss of death. Yeah, and they were the rising star. They won hot a startup. But what's your take? How do you guys different from, say, HADAP? And again, we're gonna start to see the bloom come off the roads. And a lot of these startups, you guys are clearly clearing the runway with the Series B. You're a proven entrepreneur. You've been there before. What do you guys do differently that the market needs and how do you see the market? I think it's a great question because there's so much interest on SQL and Hadoop and of course Hadoop was one of the first organizations to do that and we hear so many people getting into the space. But we're very different from all of the other players because most of these players have developed architectures for analytics only. And what that means is that they can't serve real time applications. They can't change records on the fly. They can't do transactions. And if you're going to be a general purpose database that's going to power applications on the Hadoop infrastructure, you need all of the services that traditional databases have. Databases like MySQL, Postgres, Oracle, SQL Server, these are the general purpose databases that today are hitting the wall. And now people are looking for scale out solutions and they can scale out on NoSQL but of course they give up on all the SQL services or they can scale out on the NoSQL architectures that are out there that are unproven proprietary architectures or they can scale out on Hadoop which is proven technology to work at petabyte and beyond scale and that's really what we're trying to bring people to, to power applications, real time applications and these other architectures are really about doing data science. So we had Kurt Monash on theCUBE last summer, John at the Vertica user conference and I don't argue with Kurt Monash about the database stuff anyway because it's just one of those. I mean, no matter if you're right, you'll still do it. You'll still do it. It's the point, right? You'll still bury me. But I did ask him, you know, what percent, and sort of your discourse just then reminded me, what percent of the sort of new applications really need, now you didn't say asset properties. First question, is that what you're talking about? Is it asset properties? What percent of these new apps really need those asset properties and I wonder if you could comment on that and then I'll share with you what Kurt said. I love that. I spoke to Kurt last week as well and he's a fantastic character and we're having a great dialogue. I think we're gonna be really having an ongoing communication because he really gets this space to your point. But yes, splice machine is fully acid and the question is, what use cases is that good for? Many people think that asset properties are just for traditional, transactional types of operational applications. These are the applications that people think of that may be powering a website or doing financial transactions, e-commerce and things like that but really, really transactions are required even for OLAP. If you wanna update secondary indexes in a SQL database at the same time as the data and keep that consistent, you need a transactional context. That requires asset properties. That requires a traditional architecture that allows you to keep those atomic consistent and isolated transactions in the database and that's precisely what we're seeing out there. Even if you're doing a large scale reporting application or an analytics application, you actually do need transactions. You need asset properties and that's why we built what we built. Yeah, so one of our other CUBE alums, I was going on Twitter with Ray Wang the other day and we've actually forecast the sort of SQL and no SQL spaces and there's a lot of growth in SQL. I think the Hadoop has been a tailwind for SQL. I think so and I think that people still to this day think of Hadoop more as a static data repository, something that you do data science on. They don't think of it as real-time yet, even though HBase has been around for quite some time and powers many different real-time applications, but today I think we'll start to see real-time SQL-based applications powered by systems like Spice Machine. Well, and we get discussions all the time with some of the early Hadoop practitioners like Stefan Groshup from Datamere. So that's not what Hadoop was designed for. What are you talking about, Monty? Well, I just beg to differ. What gives you confidence that you're going to sort of break that mold? Well, what gives me confidence is what our customers are doing. We have one customer who's got this tremendous direct marketing service that they're providing for retailers, large-scale retailers out there and they have implemented an off-the-shelf campaign management software package that's powered by Oracle. That happens to be a campaign management application that's sold by IBM and it's powered by Oracle. They hit the wall and they said to us, listen, we really need to scale out to make our business more profitable and to grow the way we want to grow and we're thinking about replacing the Oracle piece of this. Can you do it? And so they gave us a bunch of queries to test that have been giving them problems and we just blew it out of the water. We just were performing better on much less expensive hardware and then we connected it up with their Cognos systems and their Abinicio systems which were essentially used as their ecosystem for data and everything worked fine. And now we're integrated doing full end-to-end campaigns on their campaign management system proving out that Hadoop can power a real-time application. So that's an example for you. Monty, what was the catalyst for that example? We're talking about replacing Oracle. Was it just the sort of Oracle squeezing them and license costs? Was it the business outcome that they were desiring? They knew they couldn't get it with Oracle. I wonder if we could talk about that later. I think that, I don't know the details too much about why they turned to us but I do know a few of those pain points. Number one, they were having some performance problems in the Oracle architecture they were using today and they were looking on expanding. And I suspect Oracle was probably looking for scale up to Exadata or perhaps in other ways scaling out but I know that they were looking at how they would stay on Oracle and still serve their issues and that was quite expensive. So it's a combination of this price performance issue that they're really trying to get at. And I see that across the board in many of the companies we talked to, they hit the wall with their existing database, they're looking to find some solution, they need to scale out and they really didn't know they can scale out on Hadoop yet and with us coming in and educating that Hadoop is real-time, we're seeing some traction. Let's talk a little bit more about a splice machine, the company, where you're at with your funding and your head count and things like that, all the key metrics there. It's a fun day for us because it's a day after our Series B financing, as you mentioned earlier. So we raised $15 million from two great venture capital partners, our original investors, more David Dow ventures as well as InterWest partners. We are excited to take the company to the next stage. We're about 30 people. We're in private beta. So we're a young company still, having a number of about 15 companies testing our system at scale with their use cases, like the one I mentioned earlier. And we'll be in public beta sometime very soon. I look forward to reaching out to you guys to announce that, but we'll put our platform out on our website for anyone to try, download, and even use at small scale. So you'll see that public beta coming real soon. What are some of the success points that you're looking for out of the public beta? Public beta is actually pretty important. We have our startup at our crowd chat, which is actually public preview beta. I want testing at scale. What does that mean scale? And what are some of the economics involved? You don't have to give us the exact numbers, but kind of order of magnitude, how big a scale are you talking about both technically and then on the outcome side from the business model standpoint, what does that render into? I mean, I'll say $15 million financing. You got to show some traction. So just give us a little taste. Sure, sure. So some metrics of success for me. I'd like to go into 2015 with thousands of nodes deployed out there. I'd like to see, I would say, south of $5 million of contracting activity for a first year of a startup company having a generally available product. Anything north of that would be an impossible task, I think, but I am very excited about reaching those objectives. I think that the interest levels that we're getting in conferences like this here at Strata as well as just on our website, without even having our generally available product tells me that this is a market that's quite hot out there. And with respect to the scale of the kinds of problems that people are trying to solve, I would say they're always greater than terabytes, obviously, because otherwise traditional database technology would serve them well. But tens of terabytes and hundreds of terabytes are the sweet spot, I think, of this marketplace. And then there are a few players out there that have petabytes of problems. The tests that are being done by our charter customers, the beta customers I talked about are the kinds of things where 20 billion or 100 billion rows are being joined against a similar size table. So we're talking very significant size that would overwhelm traditional technologies. And it's remarkable at how easy it is for companies to actually generate that kind of data and make use of it today. That's fantastic. And total finance that you guys raised, A and B total, what is that coming through? It was $15 million in this round and $4 million in the first round. I'm not a fan of raising humongous amounts of money. I think that can take startup teams and defocus them if they have too much money. I think you should raise exactly what you need and that's what we need. Great. And obviously the first round, it's not that a lot of cash for that, given the scale you are, but great validation. Again, this is a market of validation. Right now we're seeing this in the big data. Monty, thanks for joining on theCUBE. I'm John Furrier with Dave Vellante. This is big data, SV, we're covering all the action, entrepreneurs, investors, big companies, big moves here in Silicon Valley, covering all the news of the Stratoconference, all the tech athletes. This is theCUBE, right back with our next guest.