 Back here at Hadoop Summit, live in Silicon Valley. This is the heart of Silicon Valley. San Jose Convention Center. I'm John Furrier, John Furrier with Jeff Frick. Here, this is the cube. I mean, Jeff Kelly, sorry, not Jeff Frick. Jeff Kelly, the big data analyst. This is the cube, our flagship program. We go out to the advanced instructor to sit down from the noise. And Yves, say your last name. Yves Doncheuil from Thailand. Hello John. I could never do that. Hi, Jeff. Welcome to the cube. Ladies and gentlemen, welcome to the cube. We've been called in a couple of minutes in preparation for this evening's reception. We are here live. They're announcing the cocktail reception. This is the cube. We're on the ground floor here at Hadoop Summit. Yves, welcome to the cube. So what do you think of the show so far? Let's talk about big data. First, what's your take on the show and then what you guys are doing here? So the show is extremely interesting to see how fast it has grown. I mean, they were saying that it's like the sixth or seventh edition. I think it has actually, I think it started more recently than that. It started to pick up more recently than that. Now, you're seeing how they have 2300 or 2500 people in this gigantic convention center that they got 60 or 70 software vendors who are actually here to present their wares and to exhibit is actually extremely interesting. Big data is probably the fastest adopted technology in the history of IT, at least in the 20 years that have been following the history of IT. And it's amazing to see the use cases and the results and the feedback that we are getting from the community in such a short time. The big theme here is obviously data. We hear data lakes, now the buzzword. I mean, I consider it more like an ocean, but lakes are nice word pooling, small data lakes. Integration is really a critical thing. And getting value out of the data is the number one thing we're hearing besides enterprise grade, which is the theme of this morning, making it enterprise grade. But the real challenge in getting the value of the data, what's your take on where the industry's at on that piece of it, getting value out of the data? Because on a straight up green field approach, someone who's starting from blank sheet of paper, it's pretty straightforward, architect a solution and get the value there. But for a big company in an enterprise who has existing stuff, legacy infrastructure, legacy software, it's challenging. What's your take on that? Well, John, that's the key really. Big data is not about inventing new stuff. It's not about inventing new data. You have your legacy, you have your system, your transactional databases, your various applications, hosted in the cloud on premises. You got all those systems which are producing vast amounts of data that have been untapped so far. You know, Mervedrian this morning was talking about dark data. And all of that stuff is the raw material that will be used to actually feed your big data systems, to feed Hadoop or other applications, and to actually extract value out of it. So a big data project doesn't grow out of nothing. It grows on top of all the existing data. And one of the keys is to feed this data leg, this data ocean, whatever we call it, to bring the data inside it from wherever it was residing before. And that's why integration is extremely critical to the success of big data project. Get the data, get to it, get it into the big data system, getting into the big data application. Now that's only the first step, to bring the data in. Then you don't want your big data system to be a new island, to be isolated in the middle of that lake or this ocean. It needs to be connected to the rest of the information system. The actionable intelligence you're going to produce out of this big data. You want it to be fed back into your operation or applications. You want to get into your analytics systems. And hence, keeping the connectivity, keeping the integration bi-directional between your big data applications and the rest of the information system is actually also extremely critical. Every CIO that I talk to in the enterprise, these are large-scale enterprises that we do a lot of discussions with. Certainly Wikibon has a community of CIOs and practitioners who are always talking about this as well. I know Jeff and Dave Vellante and their team are doing a lot of research in the area. But when I go out and ask people directly, hey, what's going on with big data? They all say, yes, we're doing it. We have a big proof of concept. Depending on the scale of the IT budget, the proof of concepts are bigger than others are smaller, depending upon what they're going to bite off and chew relative to the POC proof of concept. But the number one thing that comes up besides security, which is kind of assumed, that that's an issue to be taken care of, is compliance. It's compliance is the number one issue. And that's around the governance side of it. So what's your take on that? I mean, where are we? I mean, there's a lot to get done and that seems to be an inhibitor to adoption. Are you seeing the same thing and what are your thoughts on that area? So big data projects are starting primarily as pilots. And again, Murph this morning was talking about Chuck with his cluster in the closet or cluster in the cloud. He's just paid for it with his credit card. That's how most of the pilots are starting. And when you're doing a pilot, you don't really care about governance of information. You don't really care about compliance. So don't really care about IT policies and strategy. All you're trying to do is prove your point, prove that you can actually extract value out of those massive, untapped amounts of information. But you're absolutely right. Once you get past the pilot stage and once you want to start to deploy those projects in real life, into your enterprise IT environment, then you need to start to adhere to the policies of enterprise. And that includes the governance. And the governance, there is part security, there is part compliance, there is part traceability of the data. There's also a big issue that comes into play which is the data quality. Data quality is something that I would say reasonably easy to master. Once when you are dealing only with data that you control, that you own. But when you're starting to get into data that's coming from third parties, partner data, customer data, open data, social data, where you don't have the option to go back to the source and to actually fix quality issues, then the governance of the data becomes even more complicated than it was in a non-big data environment. So you guys have a discussion on Twitter. I saw it going on. I just checked out the article on the VarGuy, covered you guys, your partnership with Neotechnology. We've been following that technology. I've been following that technology literally going back to 2007. Facebook and they were big customer graph databases. We're just getting onto the scene. Now graph databases are big. That's this point to some of the emerging trends. So can you talk about your partnership with those guys? Absolutely. So historically there's been a clear divide between I would say the traditional data residing in relational databases and then the new sets of data, semi-structured, multi-structured, unstructured that you would just dump into your Hadoop cluster. The new databases that come out there, the no-SQL databases, the new SQL databases actually bridge the best of both worlds. They offer both the structure and the I would say degree of protection that a relational database would bring as, and that's on top of that, you are getting the flexibility that you would be getting from unstructured on semi-structured databases. So we see a lot of traction out there for adoption of those databases, especially as big data applications start to become feeders of operational systems and not only analytical applications. So we have worked with the leading vendors of no-SQL technologies. So MongoDB, Cassandra, HBase, Neo technology. We have made a few announcements recently. Couchbase is another one. The goal being of course to be able to support all of those technologies and to offer the broader choice to our customers. We don't want to constrain them into making technology choices. We want to offer them the ability to support whichever technology choices they are going to make. Yves, I want to turn to Talon, the company. Recently you've announced a new CEO. Your co-founder is now taking on more of a strategy role. Similar to the news we heard happening at Cloudera recently. Tell us a little bit about that. Why the change? Jim Foy now appointed as CEO. Why the change? What is the signal for Talon, the next part of Talon's life? Right, well Talon does reach I would say a fairly significant size. We are now approximately 400 employees in seven countries, deploy three continents. We got several thousand enterprise customers and we have been growing quite aggressively since inception. In 2012 we grew 54% compared to 2011. So we've reached a stage where we needed, I would say, more experienced people into bringing the company to the next stage. And Bertrand and Fabrice, our two co-founders, have done a tremendous job of building that company from scratch and they remain extremely involved. Fabrice is still the chief technology officer, is still driving the product strategy. And Bertrand, instead of doing, I would say, everything else, is now going to focus on strategic direction for the company, which markets we want to enter, which technology we want to invest into, and leave the, I would say, the operational management of the company to Jim Foy. Jim is a very experienced software executive. He has been managing many software companies in his life and with lots of success. So Jim is going to bring us the experience that we need to take talent to the next stage. So talk about kind of growth and moving to the next stage. It's all about growth and continuing to sustain the type of growth that we have been experiencing in the past, absolutely. So I also wanted to touch on what you're seeing in kind of the rest of the world. You're obviously a French-based company, a French-based company. What are you seeing in terms of adoption of big data technologies, I guess, among some of your European customers, versus maybe what we're seeing here in the United States? Are you seeing any differences? I think there was a perception that maybe Europe was a little bit behind the US in terms of adoption of big data. Is that accurate, or what are you seeing? So first, Jeff, I'd like to correct what you just said. We are a global company. Even though the company was born in France, I would say today we got probably half of our staff in North America. And most of our management team in North America. But it's true that we got French roots, that we are extremely proud of, and it would be difficult for me to hide them anyway. And we are extremely well deployed in other European countries, the UK, Germany, for example, as well as in the Asia-Pacific region. So going back to your question, adoption of big data. The, I would say Europe has always been, I would say, a few months behind the US in terms of technology adoption. Now, this is also probably impaired by some of the stronger regulations in terms of data privacy in terms of where data can actually be hosted for European companies. But there are a number of companies in Europe which are managing extremely large data sets. For example, telco operators or online auction platforms, which have their European branches on European operations. And those are, I would say, ripe for big data as well. It just need to check the, I would say, the European extension strategies of our partners, Rotten Works, Cloud Era, MAPPA, are all opening European offices. And if they are opening European offices, it's because there is business in Europe. Open source has always been, I would say, extremely well-received in Europe. The fact that big data and Hadoop were born as open source technologies is also clearly an adoption driver in Europe. Okay, my final question for you is, share with the folks out there about the company. What should they know about your company's value proposition? And we'll have to get the final word in here as we've got our next segment coming up about the company, company's culture, the value proposition, and the products. What is the main value proposition you guys offer your customers? Well, quite simply, it's to facilitate the adoption of big data technologies. There is a clear shortage of expertise. When it comes to big data, I mean, Hadoop remains a completely dedicated platform to adopt. It's complex to develop in MapReduce. It's complex to develop in Yarn. Learning new languages such as Pig, such as HiveQL are extremely complicated. What talent brings to the table is the ability to abstract the development of your big data technology through a very simple, graphical layer. And basically anybody who is capable of doing data integration in the traditional world can adopt big data integration and use it as easily as if it was not with new and revolutionary big data technologies. Yves, thanks for coming outside the queue. Obviously, we're not going to hold anything against you being a French company, but now global, the food's great. When I lived in France, it was a great culture of excellence, and certainly great food. So you guys certainly, and you live longer, you drink red wine, that's proven. Big data has proven that. And we also develop great software, John. Congratulations. Thanks for coming inside the queue. Great to have you. We'll be right back with our next guest after this short break. This is live from Hadoop Summit. This is theCUBE. I'm John Furrier with Jeff Kelly. We'll be right back.