 Hi, we're back. This is Dave Vellante of Wikibon.org and this is theCUBE, SiliconANGLE's continuous coverage. This is Big Data Week. We were at IOD, IBM's IOD on Monday and Tuesday. We're here at Strata Wednesday and Thursday. We're here live in New York. And this is where all the Hadoop crowd gathers, the Big Data crowd, the data scientists, the BI professionals, the new platform guys, the startups, the Big Whales trying to get a big piece of that pie. It's just, it's the place to be for Big Data. And we're here with George Matthew, who's the president and COO of Ulteryx. George is a CUBE alum. Welcome back. It's been a while. It's been a while, but good to see you again, Dave. Thanks for having me. Yeah, we were just talking about, the last time we had you on was at Sapphire. It was Sapphire 2010. Yeah. And wow, I know the CUBE's come a long ways. It has, it's quite an operation here you got. Yeah, well, thank you. And well, we've also seen the evolution of the whole Big Data business. I mean, at Sapphire 2010, as you remember, we heard, we heard about HANA, not much about Big Data. In fact, I think Bill McDermott might have mentioned it once in one of his keynotes. And then this year, maybe you heard it 20, 30, 40, 50 times. You know, SAP's more about fast data. Yes, yes. But boy, we've heard the term Big Data here a lot. As I say, the industry's evolved. Everybody's kind of redefining things. I mentioned I was at IBM IOD. IBM's super gluing its analytics business to the term Big Data. So everybody's in the business. So how you doing? Great. You know, we are right in the middle in my view of the third inning of a nine inning ball game around Big Data and analytics. And for me, I started to see how infrastructure has largely started to evolve in this space. To think about Big Data today, it's much less about what the infrastructure around Hadoop and my sequel looks like because that's actually played itself out in the first three innings of the ball game. But at the same time, people are now trying to understand what the business value is, what the outcomes are, how to make analytics more consumable, more approachable to a broader set of users. And that's what we're doing at Altrix. We see ourselves as an analytic engine that can enable people to consumerize Big Data faster and cheaper and more efficiently than anything else in the market. So you got a long history at Altrix in BI. Yes. You got started at a time, you kind of started before the Enron debacle and you know, people were putting forth this promise of 360 degrees views of the business and cubes and all that other good stuff. And it's sort of that whole industry became one of largely of reporting. So looking back in the rear view mirror and the Big Data business has completely changed that. It has. Talk about that a little bit. Yeah. So when you look at where the change in the market has occurred and you know, having my experience been at Business Objects as many years as I was and saw how a BI stack was largely about a data warehouse, ETL and business reporting, dashboarding on top of it, the challenge for the analytic space, it's not about being able to distribute a report to hundreds of thousands of users. The challenge is how do you take all of the data that you need from inside the firewall, from outside the firewall, from social media sources, from cloud sources, package and productize that in a way where you can get consumable insights out of it and share that to a broader set of users. And by the way, you want to be able to structure a solution where the single person, the data scientist, the data artist and the data analyst, they can actually be the ones that are conveying results. So the biggest difference I see between the world of BI and analytics is that BI is about reporting the past. Analytics is really about thinking how a analyst or a data scientist can convey the not too distant future and give results of what the next best decision should be as opposed to reporting for hundreds of thousands of users of what the past was. I mean, I've been maybe unfairly critical of the whole BI space because really those two things, one is it really was a rear view mirror looking activity and it was important, especially again post Enron, it became critical, but on the other hand, a lot of that was compliance and just not driving a lot of real, exciting business value. There was some, putting beer next to diapers, that was another famous example, that's cool, but rudimentary compared to what we're talking about today. And then the other piece of that, which you touched upon is really didn't scale. The adoption of BI was very much limited to those analysts that could really do things with the data and then yes, put out reports. And I had another question and I had to wait six, nine months and I had to fund it and it just wasn't the right model. Talk about how that's changing and what you're doing in terms of participating in that. Yeah, so that model's radically different today now. So Dave, when you look at this market, the analyst has responsibility for making the next big critical decision. So I'll give you some examples of this. Customers like McDonald's, when we look at how McDonald's uses analytic solutions like Altrix, they go ahead and figure out how do they decide where they place their next store based on customer propensity, demographic insights, how much point of sale data they have, the network effect of not only where they place their current stores, but where they place competitive value to the distance of other stores. And you have to pull that all into one composite view to share to a broader set of users to decide how do you make your next decision, in this case, where you open your store. And that kind of analytics is much, much different. Much, much different than what is effectively a report that looks as the past. And so what we believe is you need the ability to take the analytic processing around data wherever it comes from, be able to package predictive analytics right alongside of how you'd package data along with the predictive capabilities and publish that as a fully formulated application to share to a broader set of users. And the importance is that you got to have it be able to be done by one person, not by a team of or staff of IT managers and members to deploy a solution in, as you've stated, nine months or a year because you need that answer right away instantaneously in the hands of that end user. Interesting dynamic here, George, is that the old world is not sitting still. I mean, they've been disrupted before, they've seen this movie, and now they're really charging hard at this. How do you see the sort of old and the new? You're coming at it from a hybrid version of that. You got a lot of experience in this world. How do you see the sort of new, fresh startups that are immature, coming together with the big whales and getting guys like you, the tuna in the middle, you got the whales, the minnow, and you're the kind of the tuna between. How do you see all those coming together? Well, you know, anytime disruption happens in the technology industry, it doesn't happen where someone completely gets replaced. Effectively, what ends up happening is that new things form alongside of the old things, right? So what we're seeing is a level of substitution. Let's give an example of this. So instead of actually doing a full scale enterprise ETL process for your new applications that you might be standing up, what are you ending up doing? You're actually putting it inside of Hadoop, right? You're going ahead and using cheaper resources, being able to more effectively get analytics delivered to a broader set of users without having to go rely on the infrastructure of the past. So I think this is where you're not going to see this situation that anything that's new coming into the big data and analytics world is somehow going to replace or supplant everything that was before it. It's just that you have a new opportunity to solve problems that weren't possible before, right? If you wanted to combine, say for instance, telephony data around wireless signal strength, drop call volumes and combine that together with information that is text processed and analyzed from, say for instance, a call center environment and bring that together to compose a churn analytic, you couldn't actually do that. You had to rely on just the data that was inside of your call center to assume what the churn analysis was. But now you can actually combine sources of data together in a seamless way and suddenly you have a better churn analytic model. Well, it turns out that eight out of 10 telecommunications wireless carriers in North America are using altrix for better churn analytics because you can do that. It's not like the BI reporting in that organization has gone away. It's just simply that we are solving a use case that hasn't been solved before. Yeah, so talk some more about the customer use cases. I mean, telco's a big one because of the churn. Talk about how you've had business impact directly affecting that churn. Yeah, so churn is a big challenge for telecommunications companies and particularly wireless carriers today because there's no real reason why a member or a subscriber of a carrier can't actually go over to another carrier because the thing that blocked them before was number portability, right? You wanted to keep your number and if you decided to change your cell phone provider, you had to get a new cell phone number. But about four or five years ago, they introduced number portability to the equation. So there's no issue in terms of your ability to leave. So you have to be competitive as a wireless carrier, for instance, on two fronts. One is you have to maintain a competitive ARPU, average revenue per user, and you have to be able to give valid offers to an end user the moment there is a likelihood of churn. And that's what solutions like altrix do for our customers. We actually go ahead and help them understand what is the most probable situation where you're going to run into a segment of your customers that are going to churn out and how do you make an actionable return on that potential of churn and give a offer back to the end user so that will prevent them from going to another carrier. And so those kind of solutions where you're actually making analytics more operational, assuming that there is this answer that comes back faster than what's in your rear view mirror as far as your reporting and your dashboarding goes. This is why those people left. Yeah, this is why those people left. That's not what you want. You want to stop it before they walk out the door, Dave. And that's what I see is really an opportunity in this market that's different, frankly, than where things were five, six years ago. My colleague, David Floyer, we always have these debates about what is real-time and near-time and everything else. And he defines it very simply. Real-time is before you lose the customer. Yep, that's a good example. So you're participating in real-time in this context. Yeah, and it's actually going to get even more real-time as you go, right? Because look at what has been announced even at Strata, right? Suddenly, you've got distribution providers like CloudEra starting to take largely what was a batch-based infrastructure in Hadoop and now turning that into real-time with their Impala announcement, for instance, right? So this world is now starting to drift much more towards a highly available, highly scalable, redundant environment where you could actually get analytic insights, you can get transactional processing in real-time without having a lot of the infrastructure weight that held people back. And that's what's really incredible about the next five years in this space. So if you're a CIO with your application portfolio, let's focus on the analytics piece of the business, the data warehouse infrastructure, that piece of it. You've been investing for years. I've often said it's like a snake swallowing a basketball. Every new thing that comes out, you try it, you patchwork it together. Because you see the potential, right? And to maintain a degree of competitiveness, you have to keep investing. So you're on this treadmill. Now all this new stuff happens. So how would you advise, and I know it depends on what industry or factors, but you know, generically speaking, give us some horizontal advice that we can chew on. How would you advise they allocate that portfolio? Like an investment manager? So if I was a CIO thinking as an investment manager, what I would actually look at is the operations that keep my lights on from a data analytics standpoint, mostly residing in the data warehousing and ETL solutions that you've invested into. And to go out and rip and replace all those things just doesn't make sense. But at the same time, if you're looking at ways that you handle unstructured data that you haven't been able to take advantage of before or highly petabyte scale, sensor data that you weren't necessarily put inside a relational data warehouse, now you have an option to have all the things that are structured, unstructured, semi-structured, spatial, non-spatial, really doesn't matter, coexist very nicely. Now it just means that you are investing into, say for instance, no SQL databases now, you're investing into Hadoop infrastructure, but your current investments don't have to go away, it just means that you can stand up and test these things very quickly and efficiently without spending a lot of money, right? You're talking not like millions of dollars, you can actually get most of this stuff working with a few resources and very little money as far as the infrastructure spend goes. What I see is that these test beds are going to become much, much more prevalent for a CIO. Their ability to try things that they haven't been able to because it was too costly to do it with their enterprise infrastructure. The enterprise infrastructure is, the investment has been spent, it keeps the lights on for the current business, and what you are going to be able to do is have much, much more interesting test beds of activity, particularly around your data and analytics that you couldn't have before because of what's possible now. So in follow-up questions, this is Y2K, I mean, IT to a lot of CEOs has been a big sucking sound and we've seen IT as a percentage of revenue decline, it used to be up around seven and eight percent in some industries down to two to four percent, depending on what you're looking at. Do you see the potential, and maybe you can give some specific examples where the productivity impacts of big data analytics are going to be so profound that CEOs will actually start investing again in IT as a percentage of revenue and have a gain share? Well, one of the things that we talk and see is the decline of IT spend, right? But let's actually look at that a little more carefully because is it actually declining? I don't think it is, and here's why. What I think is actually happening right now is that the discretionary spend that a CIO has on basically capital expenditures is going down the tubes, right? Because you don't have a lot of money to put more CapEx investors other than keeping the lights on for the stuff that you've already invested in the non-discretionary side of things. At the same time, the line of business is actually using their OPEC spend to spin up all kinds of interesting things, right? And largely that's happening through SaaS, right? So when you look at where analytics is a service, right? Big data platforms like what Altrix provides as a subscription-based service to the market goes, we're not selling to the CIO. Guess who we're selling to? I'm selling to the CMO. I'm selling to the VP of churn analytics. I'm selling to the head of RF engineering. I'm selling to real estate, the VP of real estate. And the reason why I'm doing that is because they have operating expense as part of running their business. I am giving a very viable solution that is packaged as a service, right? And they're able to consume that without having a big infrastructure and CapEx spend associated with it. Yeah. So this is why I don't think IT spend has actually declined. It just so happens that the power base of who's actually spending on IT has moved from the CIO to the line of business. And if you're a CEO, you're actually pretty excited by this, right? Because you can actually then control and say, hey, by the way, is this actually having a maternal and material impact to your business and the growth or profitability of your business as opposed to just the cost of doing business. All right, George. George Mathew from Altrix. Thank you very much for sharing us with us through your perspectives. We got to go wrapping up the day here from Strada and Hadoop World. This is Dave Vellante. We'll be right back with theCUBE with our next guest right after this message. Dave, thank you again. Appreciate it. Thank you. See you. All right, George.