 from Union Square in the heart of San Francisco. It's theCUBE covering Spark Summit 2016, brought to you by Databricks and IBM. Now here are your hosts, John Walls and George Gilbert. We'll go back here on theCUBE, we continue our coverage here in San Francisco of Spark Summit 2016 along with George Gilbert. I'm John Walls and we're joined now by Steve Huber, who is the Vice President of Sales for CASC, which is an open source applications platform. And Steve, thanks for being with us here on theCUBE. Thanks for having me. First time, I believe, right? Yes, first time. You're a CUBE newbie. CUBE newbie. So tell us a little bit of introduce, if you would, CASC to our audience and primarily what you all do and what brings you here to the Spark Summit? Yeah, absolutely. So CASC is a little over four years old, we're an application company that's built an open source platform to support building applications on Hadoop and Spark. And so the ecosystem has had a lot of challenges dealing with this rapid growth of many different components in Hadoop and in the Spark ecosystem. And so we make that a lot easier to deal with and help people bring value to those things much faster. And so, how do you do it? Yeah. That's what you're all about. It's super soft. You've got a lot of data happening here and then you got to make sense of it at the same time, so. Right, so again, we have this platform called CDAP, stands for CASC Data Application Platform and it's really an abstraction layer that would sit on top of Hadoop or on top of Spark or both and enable organizations to build data pipelines, track and manage lineage and actually build full function applications and manage that from development all the way through to production. And typically our customers find an experience of about five times faster in using something like CDAP versus having to stitch all of those components together manually. So, do the Distro Vendors, the Hadoop Distro Vendors or even Databricks see you as someone who gets in between them and their customer? So, I don't know if I'd say we get between them but we're really an enabler, right? So, CDAP is an open source platform that supports all the different flavors of Hadoop and so we are a very portable solution. In fact, one of the nice things that our customers like about CDAP is you can build pipelines or applications on one distribution and you can run them on any of the distributions whether on-prem or in the cloud. So, we really sort of sit above the storage layer and if you will, in the stack and we're focused on adding value as opposed to getting between. But the distros have begun to fragment. Yes. So, you have to do either a lot more work to provide that commonality. Well, and then it's not ore but the platform vendors who are diverging all of a sudden see someone with a layer above them that makes them look more homogenous. So, that's a great point. I mean, ultimately some of the larger customers that are a little more down the maturity curve with technologies like Hadoop and with Spark actually have started to implement different distributions in the same organization. So, portability has become a really significant differentiator for someone like Cask in the marketplace. And the fact that we can also manage that full life cycle for those customers has been really important. So, it is a lot of hard work and our engineering team out here in Palo Alto does an incredible job of staying current on all the major distributions. And this portability, is that synonymous with interoperability? Absolutely. I mean, that's what we're talking about here, right? Certainly. I'm sure that everything's syncing up, talking, speaking the same language. Right. You literally could build as an example a data pipeline to support a specific use case and on today running on one distribution and tomorrow plug and play and with minor configuration change running on another. The alternative is to rebuild all those integrations from scratch. Right, so we're really helping companies to stay on that edge of innovation and be able to really be agile in their utilization and consumption of Hadoop and Spark. What about if there's some basic building blocks on the bottom in the distribution? Not you, but below that, that are just missing. Like, one distribution just might not have a, it's not lacking a consistent security model, but like any security in certain components, whereas someone else has implemented very consistently across the modules. Certainly, so one of the beauties of having a platform underneath is that you can build a lot of things very quickly. So it is really designed from the ground up as a developer platform. So when there are things missing or custom things that are unique to a specific use case, we can very quickly enable the creation of those things and fill those gaps or bring in other third party technology and write connectors to it very, very quickly. So in other words, part of your strategy explicitly includes backfilling missing functionality. That's if required, yes. Okay. I mean, just in the world as it is now with all these great inputs that we have, right? Here comes data from every which way, right? What about just the complexity in general? I mean, how are you breaking this down for clients so it becomes actionable? That's a great question. So I would say, the way I'd like to answer that is when we do have customers that are a little more mature, again, have kind of gotten their knuckles bloodied, working through their journey in this ecosystem, they're looking for that sort of abstraction layer. They're looking for ways of being able to do it better, being able to be more agile and so forth. So by being able to remove all that complexity and give this acceleration, we really can enable companies to leverage all these data assets much faster. So as an example, one of our large customers is one of the world's largest providers of data. And so one of the things that they found that was really interesting in their evaluation process is they took a team of people that were quite skilled in the ecosystem, and they took another team of people that were sort of a mediocre skill in the ecosystem, put them through three days of training on our platform, and they let them sort of do a bake off side by side to sort of see what the results would be. So we were able to take sort of these newbies and be able to really drive a result. For them, during the POC, after three days of training, they expressed a three X factor of being able to deliver the solution faster than the organic team of experts. And that's just after a couple days of training. So yes, removing that complexity, really enabling people to get that value much faster, and the distro companies like to do business with us because we help them to sort of enable consumption and growth of the underlying infrastructure much faster. So it's a bit of a damned if you do, Dan, if you don't, for the distro vendors, you help customers absorb their infrastructure faster, but you also insulate the customers from being too tied into any one distro. Yeah, I mean, certainly getting locked in is a concern for many large companies, and they want that flexibility. But I would say we do support all the different distributions, but we have some that we work a little more tightly with at this stage of our growth, especially with Cloudera as an example. So what about applications on top of your platform? Yeah, certainly. What are, where the ISV wants, you know, either backfill functionality or protection? Certainly, so we've seen some of that. You know, one of our largest customers is one of the world's largest telcos. And so some of the other companies in that ecosystem are looking at us exactly that way, where they want to be able to build industry specific applications, leveraging our platform so they can tie into their trading partners very aggressively. And then we also are building applications on top of our platform. We actually have a solutions engineering team that we recently started to build after post our series B round of funding that we completed in the fall. And that team is focused on also delivering more horizontal applications, things like customer 360 and fraud and recommendation engines that will also be included as part of the open source platform. So to give people a much quicker jumping off point to solving those problems. So how the nature of applications that you are building and just given what we do have available now, we've talked a lot this week about continuous applications and how has that been altered now for you in terms of the capabilities that you have at your disposal? So, you know, at this stage of our development, we've been, we've tried to stay very focused on where we sort of feel the mainstream of the requirements are. So I would say, if I were going to say, these were the top three areas where we've been focused is very much been around data lake creation and operationalizing those environments. We've been very much focused on enabling data science. So many organizations have wanted more of a self-service layer and easier layer for their data scientists to be able to consume their data assets. And then also focused on these sort of what I call unified customer or customer 360 gets, I think, a little overused. So a lot of emphasis out there across different industries on having a more unified perspective of their customer and trading partner interactions. That's the sweet spot of where we've seen the majority of our opportunities in the marketplace today. Well, it's working well for you. Working well for you. Steve, thanks for being with us here on theCUBE. Thank you so much. Nice to have you and look forward to seeing you again shortly down the road. Thank you. Steve Huber, the VP of Sales from CAST. And thank you again for being here. More from San Francisco Spark Summit 2016 in a moment. Cool. All right. Thanks, guys. Appreciate it.