 Okay, we're back live here at Strata Conference at Silicon Valley. We got eight minutes to feature a company that we like here. It's popped into theCUBE. Tim Morton from AccuNew. Okay, so thanks for coming in. I know we don't have a lot of time. Shake your hand there. We have another guest, but if the guest is late, we'll stay on, but want to get you some airtime. Tell us about what you guys are doing. Real time is a real big focus right now. We love real time, it's hard. A lot of people don't know what that means, so we'll talk about what your company is real quick and let's talk about the technical problem that you guys solve and the business value that it creates. Okay, great, we'll do. So I think real time has been a big focus of the conference here, and a lot of that is about making Hadoop faster. So we're not trying to make Hadoop faster. We always sit alongside. That's green fun, by the way. I think it's a number of vendors. So we focus on operational analytics and operational intelligence. So when seconds matter or minutes matter, where you're not trying to find a needle in a haystack, but where you're trying to expose the latest results from a particular set of metrics and live dashboards. So what we've done is build real time big data analytics as a framework on top of Apache Cassandra, and we'll be announcing a release for that next week. But that comes out of our experience of working with Cassandra customers where we were really seeing that they were struggling to deal with the data modeling and querying challenges and the inflexibility that that can lead to. So we've built an engine for helping people collect data from a range of different sources. We have Apache flume integration, message queue integration. You can set up the queries, which are basically like OLAP, real time OLAP queues. All that data. Cassandra's found a nice little niche, well not niche market, with real time, high transactional, mission critical value. Is that kind of what you're talking about? Yeah, indeed. Maybe last year I would have described it as a niche, but this year I think every Fortune 500 company we go to has active production Cassandra deployments, but many don't talk about them. And sometimes that's to do with Cassandra's great scale out and geo distribution capabilities, but often it's to do with real time analytics. What we're doing is we're making it much easier and faster and accelerating that process of delivering real time analytics on top of most SQL. Yeah, and we've also seen Cassandra also penetrate some of the storage paradigms with sands and using open source to actually move off proprietary drives into a much more cost effective, horizontally scalable environment. Okay, quick question for you is, what is the focus right now in terms of the open source business model for you guys? I mean, is it software? Is it all free? You guys make money on training and support in the usual ways, or you have a proprietary piece of code, or what's the business model for you guys? So the business model is as follows. So we maintain a distribution of Cassandra. Everything we do gets pushed back into open source Cassandra there. Analytics itself is a layer that sits on top of Cassandra. You can run it against vanilla Apache Cassandra. Analytics itself is closed source, but with open APIs. So let's talk about in-stream aggregation. So right now, the business models that everyone's going after, whether it's on the OLAP side or real time side is, the user experience is really critical. Intel talked about that specifically. Having the user input data is not the preferred method. Getting data out of the experiences algorithmically or automated way is a big focus. You guys look at that at all. Is Cassandra a good fit for that, or is there any specific use cases that you see in that area? So we work with customers who are doing social media analytics, who are collecting financial market data, who are collecting telemetry data in telcos and in manufacturing. And that machine-generated data, as well as human-generated data, the main sources of use that Acunu Analytics gets put to. What we found though is that traditional BI tools are not a good fit for helping people solve real-time monitoring problems. And neither are traditional monitoring tools. So we've ended up building a very configurable set of dashboards as a web-based tool that allows you to, that allows you to... I was looking for someone. Don't worry, I'm not on camera. So those dashboards allow you to get a lot of value out of the dataset very quickly and be able to understand its structure and understand how you can expose business insight from it. So let me take us through for the folks out there that are looking at the big data market who want to know in the Cassandra space that you guys are playing, what is the big enterprise value that you guys are seeing? And if you can rank that, is there a rank? And you can go, oh, one, two, three. Is there a way to stack rank that? In terms of the value propositions? Sure. So real-time insight, I think, is number one. So doing what you would traditionally have done in perhaps a transactional relational database or perhaps on a Hadoop system and actually moving that into a real-time framework where we're talking about data coming in and being available to make decisions on in seconds rather than in minutes or hours. Yeah, so let me ask you that question. So Cassandra has a lot of use cases of what I call in the new way. The old way is kind of data warehousing, business intelligence, monolithic systems. The new way is a little bit spread out, resource-based, horizontally scalable. And we saw some benchmarks from Green Plum saying that they're 100 times faster than Hadoop. A lot of questions on the benchmark, but that's kind of an old data warehousing model. So it's like, okay, Hadoop gets a nice little messaging there, and it's a lower cost data ware as a pre-existing market. How do you guys compare and contrast to those worlds? I mean, obviously you're in a different world. But that's moving over, so that data warehouse and business intelligence is now moving over into a new world. What's your point of view on that trend? Well, I think it's all very well-making, Hadoop, faster, but really that paradigm is about accelerating the needle in a haystack exploratory and analytics process. I'm going to ask you a question that I've not allowed you to see before, and I want you to find the answer to that and mine it from a very, very large set of data that's sitting there at rest. What we're working with is really allowing you to take these NoSQL technologies and embed insights into what things matter to you. If you're doing clickstream analysis, you know you care about users being retained between levels two and level three, and that fun is very important, and you need to be able to make decisions about how hard you make your social gaming in your time. It's critical infrastructure in the sense that it's matched to business process. Absolutely, so it allows you to codify that business insight into the analytics process. Yeah, and we're seeing that too, by the way. We're seeing that the instrumentation and the data collection is critical in this, if there was a book being written, and I'm sure guys are writing books now about process improvement around this new technology. So we're on the same page there, but I want to ask you with respect to codifying process, where are we in this? I mean, how early are we? How far along the track are we? I mean, you might argue we're early, early, or, you know. I think the early adopters have, I mean, I think the only. Can you peg a one to 10, 10 being a mature market? Are we minus one? Are we two? Are we one and a half? So I think in the space, we're probably about three to four now. So the early adopters are having proven successes and are beginning to talk about those successes. I think the early mainstream are beginning to recognize what some of the early adopters have done and say, hey, I want a bit of this. I was talking to Billy Bosworth at Datastacks and we're talking about Cassandra and we did the Cassandra Summit Cube there and a great insight there. And Flash has been a real big enabler for Cassandra, given some of the latency issues around storage, storage architectures. But we talked about Cassandra, it gets kind of overshadowed by Hadoop from the hype side because Hadoop is, you know, Hadoop, right? So we kind of use the NASCAR analogy. It's like Cassandra's in the track and might slingshot around and take the front position soon in terms of recognition. Not there yet, but it's different. And I want to ask you this question because there's a point there is we're hearing that, we heard it stat earlier, one in five Hadoop projects make it to production of POCs. Cassandra seems a little bit more mature in the sense of in production. Do you see the same thing? And what's your comment on that? No, I agree. And if that was Billy's insight, I think that's very true. Cassandra, because traditionally it's been a relatively heavy weight process with perhaps slightly steeper barrier to entry and learning curve than other systems, actually once people commit to it, they find that you end up getting into production in a much larger number of cases than with perhaps an average Hadoop distribution. I would go back to the race analogy. I would say it's actually, you know, it's not a single race. We're looking to address different related, related but different problems here. Okay, Tim, last question. We got to get the hook, because our next guest is here. Thanks for coming on the queue. But final question is, what's your outlook for this marketplace? The collision of top down, bottom up, action between all the no sequel and sequel environments or unstructured and structured. What's your take on the forecast? I think the collision's inevitable. We've seen a bit of it today at the conference and with the Greenblum announcement, you know, enterprise data warehousing, unstructured data warehousing in the shape of Hadoop and no sequel databases are going to collide and merge together. But I don't know where that leaves customers in terms of their ability to understand what's happening in the marketplace. It's a tricky one. What is your take on the Greenblum announcement? I mean, just what's your point of view on it? Good, bad, medium? I think it's great. I think they have a lot of great complementary technologies and clearly a lot of talent inside the Greenblum organization to be able to deliver. You've got big money behind you with EMC. They have a lot of money behind them. I think they certainly have the ability to execute and deliver on that vision. I don't know to what extent the world needs yet another Hadoop distribution, but I think hopefully the potential competitive companies there will find a way to work together for the good of the user base. Tim Morton at CUNY, thanks for coming on theCUBE. We'll be right back with our next guest, Roger from O'Reilly. We're going to talk about applied big data in our next segment live here inside theCUBE at Strata. SiliconANGLE.com's exclusive coverage of O'Reilly's Strata conference in Silicon Valley. We'll be right back with our next guest.