 Hi, this is Stu Miniman. Thank you for joining us for Cube Conversations. Here with Jeff Kelly and the Wikibon headquarters in Marlboro, Massachusetts, Jeff released the third annual Wikibon Big Data Market forecast earlier this year and we're gonna get an update today on the Big Data Application Market. So Jeff, it said that Big Data Applications are really gonna be where the most value is gonna be created in this wave of technology. Can you give us a state of the Big Data Application Market? Sure, and Stu, great to be here. So, the Big Data Application segment of the market, which we sized recently, came out to about $1.6 billion. You think that's a pretty big number, and it is. But the reality is the majority of that revenue comes from business intelligence applications, data mining, software, and more horizontal type tools that are used to visualize, do some analysis of data, and less around the prepackaged Big Data Applications that leverage Big Data Analytics, kind of target specific business problems. So, generally speaking, the Big Data Application Market is still pretty small, and absolutely I agree with you, that's where a lot of the value's gonna be because that's where a business user, for instance, is gonna use these applications to solve a specific problem with all these insights that we're getting from the data crunching underneath the covers. So, Jeff, I remember looking at one of the reports you had done a year or so ago, and it seemed that custom was still the kind of primary application for Big Data. What really makes it so difficult to develop these Big Data Applications? Yeah, well, they're much different than your typical enterprise application. Your typical enterprise application is dealing with usually one data source and one underlying type of data. So, with Big Data Applications, you've got, from a technology perspective, the challenges you've got, new types of technology you've gotta understand, like Kadoop and NoSQL, and the different types of data sources you've gotta bring in, almost by definition, Big Data Analytics takes into account multiple sources of data. So, that's one challenge. You've gotta have that expertise. The other is you've gotta understand data science, machine learning, and all the really hard analytics work that you've gotta do. Which, again, as we've covered at Wikibon, there's a real lack of data scientists out there and people with those types of skills to do that hard science, data science. And then, of course, the other thing is domain knowledge. So, you wanna tackle specific business problems. That's how you're gonna really get the value from these applications. So, if you're doing a fraud detection application in financial services, for example, you've gotta have that domain knowledge to know what you're looking for, to speak the language, to know what the data sources are, those kinds of things. So, to build these applications, successfully, you've gotta merge those three things, the technology, the science, and the domain expertise, and that's just a challenge to do. Okay, so Jeff, with all those challenges that you laid out there, are you optimistic that we are close to seeing a broad expansion of the big data application market? I am, for a few reasons. One, I think the underlying infrastructure that's required before you can start building applications on top of that infrastructure is starting to harden and really mature. So, we've seen things like last year specifically. We saw the introduction of Yarn, yet another resource negotiator for Hadoop, which really allows Hadoop to be a multi-application framework. So, no longer is it just kind of this batch analytics platform. You can do streaming analytics. You can do graph analytics and do more machine learning. So, any number of new types of applications can now be built on top of Hadoop. And that was a real stumbling block, I think, in terms of building applications on top of that framework. Some of the other things, of course, we hear a lot about the studies that the CMO is increasingly taking over, basically, budget and spending around technology. At some point, some even suspect that at some point the CIO is gonna lose out to the CMO in terms of the biggest spender on technology. The reason I cite this, I think that's a good example, or a good leading indicator, I might say, that we're gonna see more focus on these big data applications because CMOs, marketing types, don't really care about a lot of the underlying technology. They're not gonna be spending these dollars on databases and things like Hadoop. They're gonna spend it on applications that help them solve business problems. So, we all know that the vendors go where the money is. So, I think that that's gonna help spur some of the adoption, or I should say creation of these big data applications. And then the other thing is just general maturity of the market. We've gotten to the point now where we're seeing really smart people from early adopters and really the pioneers in big data analytics, like people from Facebook and LinkedIn and even people from places like Bank of America start to leave those organizations, start their own, take their expertise, start their own companies focused on building these applications. So, it's very early, but I am optimistic. And I think we've declared in the past and we're not the only ones guilty of this, that this year or that year is the year of, in this case, big data applications. I'm not ready to declare that for 2014, but I am encouraged by some of these signs that I just cited. Okay, great. So, good to see that you're making progress. Jeff, I know you're talking to all the players out there, talking to the practitioners, the data scientists. You know, what's really catching your eye out there? Who would you say are some of the key examples that we should look at in the big data application space today? Well, as I said, it's still very early, but you know, just to hear Wikibon being an analyst covering this market, I'm lucky enough to talk to a lot of entrepreneurs who are starting new companies. One company that kind of caught my attention was a company called Wise.io. And these are some really smart guys from Cal Berkeley, astrophysicists, these are, you know, PhDs, these are the rocket scientists, if you will, who are working on solving some of these problems. And they're building applications and get very early, they really haven't come to market with their applications yet, but they're building applications that, in the case of Wise.io, bring in different data sources from a lot of the SaaS applications that people use for CRM like Salesforce or other things, and then leveraging all that data to create yet new insights that marketers or other line of business people can use. So that's one company. Another that we've talked about on theCUBE before, of course, is Trasada, Avi Menitz company, he's former Bank of America, and he's working on applications that leverage Hadoop and some of the underlying and related technologies to do, tackle things like fraud detection for financial services firms, marketing optimization, things like that. So those are a couple companies from the startup scene that I think are pretty interesting. But what I'm also encouraged by and what I think there's a huge opportunity for some of the legacy vendors, if you will, when you think about enterprise applications, a lot of people think SAP. With the introduction of HANA, they've started to build out some new applications around, specifically leveraging multiple data sources. For the most part, however, and focusing more on fast data, if you will, real-time analytics, things like a recent application they debuted was a genomic analyzer to help track and analyze data throughout the genomic research process. Another company that's really interesting, and you don't necessarily think of as an application company, you think of as a data warehouse company, is Teradata. So Teradata is starting to really put a lot of muscle behind their marketing applications that leverage all the work that Teradata has done over the years in data warehousing, but now bringing in other sources of data, leveraging some of their partnerships with companies like Hortonworks, to bring in and build applications for marketers. And this kind of relates to what I mentioned earlier that the CMO is really starting to drive a lot of purchases. And I think companies like Teradata recognize that. So they're starting to put some of their efforts into applications that sit on top of the data layer that are, you use analytics underneath to kind of drive insights, but are aimed at marketers and more business type of people to solve specific problems. Yeah, so Jeff, it's interesting. In the cloud space that I watched, there's definitely that battle for developers. And I got one last question for you. When I look at some of the platforms that are being developed, you know, the platform as a service pass has really seemed to kind of gain some increased momentum since kind of pivotal spun out from EMC and VMware. They've got a big partnership with GE doing with IBM who announced Blue Mix at IBM Pulse we were at a couple of weeks. You've got Red Hat doing OpenShift. Do you see an intersection between some of those kind of the DevOps cloud guys and kind of this big data application space seems to be some similarities into going after the developers there? I think there's definitely going to be an overlap. You know, when you talk about big data applications, you've definitely got to take that DevOps approach because you constantly need to evolve and adapt the applications. As new data sources come in, new types of analytics you want to look at. So that DevOps approach I think does fit with big data application development. You know, I think it's still very early and we'll see to the extent that it would be very interesting to see, for example, kind of the relationship between DevOps and data scientists and how they work together. Because one or the other I think challenges to building these big data applications that we were talking about earlier is that you've got to marry all that science, as I said, with actual application developments. You've got a team of really smart data scientists, for example, find some great insights in mountains of data and now you want to productionize that in an application. Well, that's where the data scientists and the application developer need to come together. And that's an area that's, I think, underexplored at this point and definitely something worth watching. Okay, well, Jeff, thank you so much for the update on the big data applications. If you're not familiar with all Wikibon's information, go to wikibon.org. If you go to wikibon.org slash big data, you'll find all of the market reports, analysis, and that's always free. Check out theCUBE, go to siliconangle.tv to see recent and upcoming events and be sure to check the SiliconANGLE YouTube channel for future CUBE conversations. This is Stu Miniman with Jeff Kelly from the Wikibon headquarters in Marlboro, Mass. Thanks for joining us.