 From Las Vegas, expecting the signal from the noise. It's theCUBE covering InterConnect 2016. Brought to you by IBM. Now your host, John Furrier and Dave Vellante. Okay, welcome back everyone. We are live here in Las Vegas for exclusive coverage of IBM InterConnect 2016. This is theCUBE Silicon Angles flagship program. We go out to the events and extract the signal from the noise. I'm John Furrier, my co-host Dave Vellante. Our next guest is Larry Weber, program director, analytics platform cloud, data services at IBM. I got the long tiles at IBM. Basically you're doing the database as a service in the cloud. Welcome to theCUBE. Thank you so much, thanks for having me. So databases as a service are a hot thing. Cloud enables a lot of stuff. Developers can plug in. All that value proposition being kicked around. At the end of the day, what is the core value proposition for your group? Because at the end of the day, I need horizontally scalable database. I need database access everywhere. I need auto scaling. The data's the key aspect of that. Yeah, so like data's that single thread across everything, right? And you think back to the days when you're talking about big data. Oh my gosh, how do we handle all this stuff? Now we're at a state in which we have all these technologies and everyone's attacking big data. And now it gets down to the point of, oh, well, I want to do that now. And I want to do it fast. And I think back to like, I have little boys, right? And watching TV, when I was a kid growing up, we used to wait for your show to come on. It was like, hey, six o'clock, cartoon comes on our TV show. Now it's like, daddy, my show's always on. I can press my button, get it whenever I need it. And that's kind of the same way now with people that are building out programs and applications. They can't wait for someone to get on the phone and bring them in and oh, let's like think through this. I want it now. I want to build now immediately. And that's what Cloudy to Services allow these folks to do. They want to essentially make, they want the infrastructure's code mandate, which is essentially been the DevOps ethos. Making that happen is two different things. Take us through what that means, because now the database is the critical aspect. It used to be also in the old days when we coded was the schema drove everything. Right now you got unstructured data, big part of it. Certainly got to run some structured data on it. Talk about that balance. You have a critical linchpin called the databases, or data, now that's going to impact other things. Yeah, so it's almost like the floodgates are open because now we have so many different data stores, data types, things that are going on out there. And you think about it back when we had big data, think about Hadoop instance. We're like, oh, semi-structured data, how do we handle this? How does that go into a database or a data warehouse? And so that problem was there, we worked through that, and then the Hadoop world opened up a little bit more. Now with data lakes, data marts, and everything in between. Data oceans. Yeah, right? But one of the other things that started popping up, and this is, it gets back to the whole like, I don't develop a revolution, so to speak. Everyone started the new king makers, all that goodness. And queen makers. Yeah, right. One of the things is people are building up their own. They're not depending the organizations that say, hey, you must use this system, this database, this data server, Hadoop, whatever. They're either writing their own, building their own, or joining the open source communities and helping build other areas up. No SQL is one of those, right? And so now we have all these different data stores across the globe, everywhere, at one point in time, that people want to work with, leverage, and utilize in their programming, and to build new applications. And from getting back to cloud data services, is when we talk about open for data and being there 24 seven, is we're also leveraging and embracing this open source technology. And we're building up and around that as well. So we're creating this entire 24 by seven, I would say storefront, for anyone, anytime, anywhere, to build their data sets, to come in and build upon cloud data services. So can you talk a little bit, Larry, about one of the things that we hear from organizations, yours and others, is you're trying to use data and analytics to help brands affect an outcome in near real time. Maybe it's before they lose the customer, or maybe it's to save a life, whatever it is. And part of that is bringing analytics and transaction systems together. Are you seeing that as a real trend today? I mean, IBM talks about it, talks about it, certainly in the Z mainframe, was one of those sort of emphases, you hear it at these different big data shows. Is it actually happening? Is it something that you guys are trying to affect? Or is it really still this world of sort of separate types of approaches? Or between data and analytics kind of approach? Well, but specifically between analytics and transaction data coming together. Yeah, so what we're seeing first off is a lot more of the analytic, you're saying if you want to capture customers before they leave, right? Customer churn models, et cetera. And you don't want to wait until tomorrow to have that happen. Be like, oh wait, you know, we figured out this customer may or may not leave, but they've already left. A lot of that's doing, one of the quicker things we're doing is we're pushing analytics closer to the data store, the database, whether it be transactional or whatnot. For example, DashDB, one of our in cloud data warehouses, we have integrated embedded analytics in that data warehouse. So you can do the analytics there, and to pull it off, you do it natively there in the data warehouse, and that's UberQuick, right? And that gives you a time of value that's unprecedented. We don't have anything else like that. So you're talking about operationalizing analytics in that example. So here's the use case. I mean, it's pretty common. I got all those disparate data sources. I want to bring them in to some platform. I want to blend them. I want to clean them. Maybe I store it in my data lake, but then I want to do a schema and read. So okay, so I have all this data. I want to get it to a point that I trust, and then I want to embed the analytics into my processes. So are you seeing that as a major trend, and how are you supporting that? Yeah, so we're seeing a lot of different things there, and I'd say how we're supporting it is, depending on the customer and what they want to do, one of the great things of having so many different technologies that we're able to handle each one of those situations. I've seen both fronts, and we're talking from where we're doing the analytics is, I've seen people do it physically within that data platform itself, but the actually be is that taking off, but also looking at it from the aspect of distributed analytics across the organization, and doing it locally, where they want to do the analytics themselves, whether it be a lot of companies are located around the globe, and they're going to pull that data in and do it where they do it locally. So I mean, it's clear that IBM has the technologies to do what I described. What I sometimes struggle with understanding is, is there a platform to actually perform that sort of process and support that operationalizing analytics, or is it more a menu of services that I'm going to pull together? How does it actually turn into a business capability? Yeah, so there's a number of different fronts. So you think about a cloud data services as ourselves, our data platform, and we think about all these different services that you just can go tap into, play with, load data into, integrated to as well, in which you can talk back and forth. But then you also have the idea of the hybrid approach that people have a lot of on-premise systems already. And so we're getting into discussion a little bit of the cloud and on-premise. And one of the things is being able to make that connection so that I can do my processing locally and then reach out to the cloud as needed, do processing where processing is done best, whether it be on a web app or a mobile app or whenever it be. And I can do the analytics there but also can bring it in as well. It all depends on the situation when the company wants. I would say it's a hybrid type platform. What's your take on the customer impact? Because this is the day, a lot of the playbooks aren't baked out on this area because everyone's really transforming in real-time. A lot of things happening fast. I need it now, I need it now. What's the impact of the customer? IBM's customer, they're out there scratching their head. They might have a problem statement that says, hey, we're going to transform our retail app. But that might not be about their software, omni-channel, normal stuff. It might be a DevOps problem. So they have no playbook. So this is a big challenge right now. What's your thoughts on that and what do you see people doing to get the data, to get the answers in the communities? Is it GitHub? Is it in some of the Stack Overflow areas? Yep, yep, where is, how does the practitioner get what they need? Yeah, so I mean, those are all great options and that's where, you know, the facto standards, folks are going there to look and ask and that's, look, this kind of supports the idea of the open platform, the open source community which people are sharing information. It's apparently different than I would say 15 years ago. People are out there offering code and saying let's work together, let's solve this problem. So that's a part of it. There's a level of education that's come together here specifically in cloud data. Now, looking at how people, how this is transforming and where people are going, I can give you an example I had just downstairs earlier on in the Expo Center, working with some of our developer advocates. These are kind of our dev ninjas that we have working for us. And inherently different than the old days was one of the apps wasn't working right on the platform on the Expo floor and it was, oh my gosh, how do we fix this? He goes, hold on Larry, no worries. Give me like five, 10 minutes. Like what are you talking about? I'll fix it. All right, whatever, go do it. He uses our own software to go create a new web app so that a user can come in there and look at the different demos that we have on the backdrop. He was like no big deal, no problem. And he's doing- He wrote an overlay app on top of the existing Lego blocks he was working with. He was leveraging all of that goodness there in BlueMix and Cloudint and actually got it up there. And if you go down there to the Expo floor now, right now that front end on those iPads is that. So let's take that step further. So let's just take that example. That's a great example of leverage. Someone already did some stuff. It's like building a house and just having someone pour the concrete for you. You don't need to be in the concrete business to do that foundation. Same with programming, whatever analogy you want to use. What do you guys have that is leverageable that the developers are really flocking towards? Is it the tooling, is it the platform in your area around the data-db? What's that key enabler that they don't have to be in the business of doing? Yeah, so a lot of it. Number one is different. You're going across the board between developers and IT architects and data scientists. And when we support them all, and real quick, if to think about it this way, in organizations that are much more mature, you might have I'm a defined as an IT architect or I'm defined as a data scientist. But in the new up-and-coming organizations, I might be all of them, okay? And so you need to be able to service and talk to and handle anything from data science all the way through, you know, app dev stuff, right, as your fingertips. Now, looking at the new models and what we're doing and why people are flocking towards looking at us, I'd say there's a number of things, right? This is a whole idea around open for data, right? Number one is we're leveraging and working with open source. We as an organization are out there, you don't have to go off one off, like roll up your own open source instance. We've done that for you. We have it there as a service. And geez, I'm trying to think how many of our 25 services are based on or working with open source. It's insane. So all of that goodness that people are building up and GitHub and these communities and learning on, we're supporting that and we're driving that, we're hosting that in the cloud. Now, what we do there is we augment that with certain IBM capabilities and technologies. Let's say before, dash DB, that's based on our DB2 blue columnar data store, but also has the teaser analytics baked in. What we're doing is we're taking what open was done and we're augmenting it with our special secret sauce and we're giving someone a unique value proposition in the cloud's data software space. So you support it, you have a foundation of open source, you take the secret sauce that's unique and that's what you charge for. That's your unique value, that's what people will pay for. In some ways, yeah. So I mean, I think it started a long time ago, we had big insights for Hadoop, right? That took our Hadoop open source, then we added some things that were IBM specific like text analytics and other things around that. And that was a while back. Now that's in the cloud as a service and this is one piece of many. I think the other aspect is that we offer so many different data sources and data stores that are integrated and composable across this that you can come in there, like compose enterprise and say, I don't know which one I'm gonna need. I'm gonna need maybe more than one database. Here you can as a developer, an enterprise developer, have it all at a glance, like a buffet almost. I can pick and choose what I want and I don't need to worry about making that decision early on. I can do it later and I know IBM's gonna support me in that process. So thinking about Hadoop, we've mentioned it a couple of times and thinking about your initiatives around Spark, what did you learn from Hadoop? I mean, IBM has its own Hadoop distribution. Okay, it's one of many. Spark comes along, you guys are all in, putting the chips on the table, so to speak. What were your learnings from Hadoop and what are the objectives with regard to your Spark initiative? Yeah, and so I might be a little skewed here because I came up through some of the Hadoop and got really hands-on early on. I'd say one of the things with Hadoop, and this might be the whole world of big data, was that, hey, a lot of noise and not people doing, right? A lot of people watching and listening and waiting. I think what we learned from Hadoop was really watching what our customers were doing with it and really looking at those unique use cases, it might not have been what we thought it was, right? It was really actually looking at and listening to our customers saying, okay, you're putting that in a cold storage, you're leaving it there, and now you're going to look at it rather. Rather pile, they're just piling up the data. That was one aspect of it. There's many different use cases here and how we can leverage it. Semi-structured data stuff that they don't know how to handle it yet, so put it out there and then start testing it, playing with it, and then, oh, wait a second, that's not a cheap data warehouse, that augments our data warehouse. We can now search that unsemi-structured data and now bring it in. It's that interplay. Yeah. That interplay really is the key. It's not a replacement. I mean, Cloudera tried to be the data warehouse of the future, but then it's all driven by the customer's objective. Beauty's in the eye of the beholder, and that's kind of the way it works, right? I mean, so take those learnings, and so you learned a lot from listening to customers and now apply it to Spark. I mean, how are you approaching Spark differently? So, how we're not doing it differently, we are listening to customers and we are seeing how people are moving about and playing with Spark. Spark opens up a lot of doors and it's different, right? It's not the same as to do, but I almost view it sometimes as almost that super force alien that sits on top of Hadoop and powers it, right? It makes it super fast, super strong. Almost you think about it like a cyborg would be. That's one aspect to it, but what I'm seeing is a lot of people using Spark to investigate new things that they couldn't do before. And a lot of, you know, with notebooks, et cetera, that people are getting hands on and they're testing out new trends that Hadoop really wasn't powered to do, and now they're testing out Spark and looking at new directions and from an IBM perspective, we are listening and we're working with them to find out those next steps. I mean, anything from, you know, leveraging, you can come down to our demo floor, it's sentiment analysis that we're leveraging Spark for, doing better things faster and empowering these customers and consumers faster. Larry, thanks for coming on theCUBE. Really appreciate it. The data is a big part of the value proposition. It's kind of like the underbelly that no one's talking about because that's where the action is and it's going to impact everything. So the developer piece is huge. IT Architect, again, they're looking for the playbook. So we'll keep an eye on that. Thanks for sharing. But I'll give you the final word. What's the vibe of the show? For the folks aren't here, who aren't on site, who aren't seeing the interactions, what's the vibe here? Yeah, so I get back to the whole thing, the theme around open for data, right? I mean, it sounds pretty simple, but it's true. It's really around taking this thread of data, it's open, 24 by seven, it's open, open source, it's integrated, it's there, whether it be the data stores itself or the data you're pulling in. We need to liberate this for everyone so that everyone can tap in anytime that they want and leverage this awesome platform of services. It's the new builders that are building on top of this and building the next generation of apps. The new world's open, it's collaborative, it's all being done in the open with people, whether that's software or content. Check out theCUBE on SiliconANGLE.tv. Go to YouTube.com, so SiliconANGLE, all these videos will be there. And of course, check out SiliconANGLE.com. This is theCUBE. We'll be right back more from live coverage of IBM Interconnect 2016 after this short break.