 Live from New York, it's theCUBE covering Big Data NYC 2015. Brought to you by Hortonworks, IBM, EMC, and Pivotal. Now your host, Dave Vellante and George Gilbert. Back to the Big Apple everybody. This is theCUBE, we're here live at Big Data NYC as part of Strata and Dadoop World. Larry Weber is here, he's the director of Cloud Services at IBM. Larry, great to see you, welcome to theCUBE. Hey, thanks for having me. So, we were talking a little off camera, but this is sort of a new part of IBM. The capability within Piziano's analytics group, it's not part of the cloud group, it's part of the analytics group, that's why you're here at Strata and Dadoop World. So, give us the lay of the land. Where do you fit within the IBM organization? Yeah, sure, so Cloud Data Services, as it is, was created earlier this year, very beginning of the year, I think I came on board to that team around March. But it's something that started, I think, last year. You think about, hey, we acquired a company called Clouded early in the year, and around the same time, we actually came out with Bluemix. And Bluemix is a platform as a service, came out with a handful of services in it, and it was around our CEO basically saying, hey, we're gonna do Cloud and do it real serious. And since that period of time, it's been about a year, Bluemix has exploded with different services, different things that are in it, whether it play with for developers, anything from internet of things all the way through Watson itself, different services are put there. At the beginning of this year, though, was decided that from a data perspective, we were gonna have our own division wrapped around Cloud Data Services. So, composable data services, and not just a database, but it could be anything from Cloud Int, could be DashDB, could be Big Insights or Hadoop, even if Spark is a beta in there. And so, what we're doing here, it's almost like the Wild West, right? We're doing it around the developers and the people that are leveraging this. It's inherently different than, let's say, back in the day, you go in to talk to somebody, hey, here's a PowerPoint, we're gonna deliver this over the next, I don't know, 10 months or whatever it is, we need a lot of integrators. Here, I can actually go out, and it was last week even, hey, check this out, look at this, oh wait, hold on, I'm gonna show you. I can go in, show ingesting Twitter feeds, live into the database, show the results. People can get started right away with this stuff, which is inherently different than some of the old legacy systems that we've had. So the mission is to package these composable services and then how do you go to market? Yeah, so most of it is around Bluemix right now, as a delivery platform for Cloudy services. But it's really looking at the developer, okay? The developer and the builder. Cloudint, when it actually came around, had a great, great legacy around helping developers and helping them build new apps and the latest great technologies based on their, you know, Cloud is a JSON data store, build the greatest apps on that. And what we've done is we've actually expanded that portfolio to say, well, there's a lot of different data types people need. It's not just Cloudint or JSON, it's not just a data warehouse. Could be Hadoop, could be Spark, could be Streams for all I know. We really want to be able to access and allow these developers and builders, give them the tools, go create the next company, create the next program, the next application, and do it today. And I think a lot of that is around, how we go to market too is, one thing that was great that came around from Cloudint was we have developers in IBM and engineers, okay? We tap them, they build awesome, great technology. Cloudint also has a group called Dev Advocates or developer advocacy. And what they do, they're kind of like, I'd say like data ninjas based on their domain. Their job is to study the data, look what customers are asking for, help customers out and just build solutions based on it. And then even open source it. You have trouble getting like some third party data, whether it be from Salesforce. We actually can actually, the developers will actually develop our advocates will go out there and build and grow and create the structure and then give it away and say, hey, anyone wants to access this, go at it. And it's phenomenal. It's a different way that we're actually helping out the environment, build out new applications. Yeah, it's interesting, IBM announced Bluemix, I guess that was early last year. Is that right, April of last year maybe? And you kind of put it out there. Hey, we're here. Yep. And it was a little unclear to me anyway, what was going to happen and you're right, it's just exploded. I mean, it's become the sort of tip of the spear for development within the IBM ecosystem. Where are you seeing the uptake? You mentioned everything, IoT, streaming apps, real time. Are you seeing any patterns developed? Well, let's think, there's been a couple, right? People like to kick the tires. And so a lot of the hot topics be like Watson, IoT, people will go look around and play with it. I've seen two different fronts here, specifically to the data services. So actually there's a number of use cases around data warehousing, but I've seen a lot of people come in and say, all right, I've got an existing data warehouse. I need to do something today. I don't have access to all those repositories right now. I have to ask permission or hey, I've hit my limit, but I want to build something or I want to try something out, even just throw some new reports up. And this is where we see a lot of traction, right? Dave, just roll up the cloud, monthly service. I can now take data off, play with it. Oh wait, wait a second, awesome. And so that's a big one. It's really the time more than the cost. This allows, enables different people in the organization to get answers faster. So what about, like you were saying, and Dave was saying, you know, Internet of Things, other use cases, are you designing it to start with a certain set of use cases where you say this is our sweet spot and then, you know, we'll leave these others for second priority, or are you following the market and following developer interest? I think both, okay, and I think early on it was a little bit of guesswork. These are kind of what we're hearing. And then over the past year, you've seen the market kind of dictate what we come out with. The best way to think of it, and some of the ways it's developed was, I tell the story of Two Little Boys, and you know, one of the things I love in Hater Legos, right? It's like you wake up in the morning and it's like you step on Legos. Step on it. Right? And it's one of those things that these are like building blocks. Okay, they truly are these services, but then again, like Legos sometimes have the kits. These overly expensive cars and you get the middle manual instruction for it. We have boiler plates or blueprints to do things like Internet of Things that would then allow different services in there. And then you actually pull it up, read it, say, oh, this is how I get started. This is where I begin. Tell us about some of these blueprints. And then also tell us about some of the composable service. Give us a little more detail on those and maybe dive a little deeper on DashDB. Yeah, sure. So some of the blueprints and their stuff, and one of the things with the Bluemix that kind of gets me is you can look today and then look next week and there's new stuff already there. It's not just limited to just IBM technology. We have third-party technologies in there as well. It's a bigger ecosystem. But there are things like Internet of Things, is definitely one of them. Leverages clouding as no sequel. Hey, you can drop off your smart device, put it on the clouding. You have, oh, with the partnership with Apple. Hey, I want to develop an app for an Apple phone. That starter kit is already in there. You open it up, it tells you what services you're going to need. And by the way, most of this stuff is free to play with. So you're like, ah, let's kick the tires. Let's start out something new. Let's develop a new app. And then, whoa, wait a second, this is really taken off. Now let's pay, let's invest. There's, jeez, there's other development starter kits, different language types that it supports. Trying to think of a few others. Oh, one that's fun. It's not, I don't know if it's a starter kit for it, but it's Twitter. We have the Twitter partnership. We take Twitter data and you can actually load that live into any of the data services you want. So, okay. So sorry, I didn't interrupt this. So in that example, a developer gets access to the Twitter firehose through you guys. Is that right? Yeah, it's actually, it's a service called Insights for Twitter, if I remember. And it's free up to a certain amount of tweets. But yeah, you can actually go, I could roll up right now, you know, call up, hey, I'm going to do a tweet on the cube and see how many people have been tweeting about that, pull that into, let's say, DashDB and then immediately look at it there and I could throw, you know, what's Watson analytics or whatever it may be to look at that, you know, study that trend or whatever's coming in off the Twitter feed. So, for those who have grown up with IBM and package software and who know about DB2 and Matiza and the blue extensions, tell us where this fits and how it's different from package software. So in other words, relate these services to those of us who are accustomed to, you know, traditional on-premise package software and how it comes together and gets delivered. Sure, sure. So these are managed services. So they're running in the background, in the cloud. You don't have to mess with them. As with regards to DB2 and Matiza, I mean, one of our great products is the protocol, DashDB. DashDB came out very into last year. We recently relaunched it with its MPP capabilities. But DashDB itself is the columnar blue technology that came with DB2, right? We took that, threw it into the cloud but we didn't leave it there. There was a blue for cloud for a period of time that we had out there. But we took that, we said, wait a second, let's do something different. We have Matiza and one of the great things is that they have integrated analytics. And Matiza box, you throw data at it, you're doing analytics physically within the box, super fast. Let's take that and then marry that with blue. And then we've done that in DashDB. So we have our capabilities. We have the full on Matiza analytics inside of that cluster that's in the cloud. And so another thing around DashDB is, and this gets back to a lot of the integrated services and composable services, when we acquired cloud and one of the first things we did was we said, wait, we have all these customers that are creating this big data around JSON documents, massive stores of JSON documents. How do you analyze that, right? It's not the easiest thing to then extract that data and put it into a database somewhere. That's like one of the hottest topics around now, which is don't just give me some JavaScript interface, but how can I get at it with the simplicity more of a SQL database? Yeah, so what we've done here is early on we built integration natively with DashDB from cloud. So if you have data stored in clouded, massive amounts of JSON documents, essentially press a button, all of a sudden creates the schemas itself within DashDB, and then all of a sudden you have a data warehouse of all that data that's there, and you can analyze it now, today. Throw some reports at it, and you've got that, wow, goodness. What is really cool about this is synchronous. So as more data is going back into clouded, it's now gonna update and populate what we have there in the data warehouse. And so that's just one story of this idea of composable services and integration across the data source. So why did you do DashDB? What was the initial motivation? The initial motivation? Oh, geez, now you're getting into history here. So there was a project back in the day, and this is before my involvement with it, was a DB2 for cloud. And it was essentially taking DB2, putting it in the cloud, and then taking that blue technology and then looking at it in the cloud as well. A lot of it was hosted in different sites outside of Bluemix, but now we're leveraging the entire Bluemix infrastructure and the ability now to speak and create with other... And so essentially you wanted to expand the scope, not beyond just DB2, right? You had all these other technologies that... Yeah, and that's why we rebranded it too. That's why we went with DashDB, because it's our own name for it. It's not necessarily DB2, even though it's fully compatible with it, because the underpinnings are. Oh, and another good thing around that, we do have native Oracle compatibility, because we've taken that with DB2 as well, as well as Natesa compatibility, to an extent. The TLCable capability. So that, the content from there. So yeah, I mean, you've got a lot of stuff going on. DB2, blue acceleration, running, you know, little end in on power eight. I mean, that's like the perfect big data infrastructure, right? I mean, because you're going to have advantages in that, you know, with your pipelining and other capabilities to run super fast analytics. Yeah, I've seen, there's been a crazy different number of use cases around this, right? I said like one of the, probably the number one one was, hey, I want to do something today. Where do I get it? How do I, I don't have the funding. Let's go out to the cloud. Let's roll up an instance, let's play with it. I also see a data science, data scientist stores in which, hey, I want to do some deep computing. This is a perfect platform to do that, because we have the capabilities in there. You load data into dash DB. Hey, ours right there. So if you know your R scripts, bam, bam, bam, you can check that out. And when you say cloud, what do you mean cloud? Is it, we're talking soft layer, we're talking other clouds? Sorry, cloud in. So just, you know, cloud in general, and you're running, you know, in the cloud. Yeah, so I'm talking specifically, this is running in Bluemix, okay? But you know, the whole ecosystem of the cloud as well. Soft layer allows like different software components and different data sources to be pulled in as well. But the managed services are in Bluemix. But can Bluemix run on non-soft layer clouds? No, best of my knowledge, no. I'm not a Bluemix die-hard. Steve Mills said it best one time. Dave, the brand is IBM. He was clear. Okay, but so, but anyway, I mean it's optimized anyway for Bluemix and soft layer. I mean that's IBM's play, you know, whether or not, I mean, somebody in IBM can support other clouds, I'm sure, our global services and so forth. Sure, sure, yeah, this is dedicated to, at least our portfolio there, specifically DashDB, is going to be in Bluemix itself. I mean, cloud in even has like a local version that you can do on-prem if you want. And so like there's others as well that you can roll up. So one of the things, one of our guys, John Furrier had dinner with Andy Jassy from Amazon Web Services a month ago or so. And he said that right now everything is about databases and Redshift is taking off unlike any other service. And the observation of one of our analysts was once you get all that data native to your cloud, then all the applications follow. Now it sounds like you've got a very cohesive story around databases. Redshift is, you know, Redshift doesn't have, as far as I understand, all the deep analytic capabilities of Netiza. It's really sort of a port of the PowerXL technology. So it seems like you have an opportunity now to really start including or drawing in data from a wide variety of scenarios that once the data is there, people build very sophisticated applications on. Is that sort of the scenario or the vision? Yeah, I would even back up a little bit and say one of the key things that we have as an advantage is that we do have all this on-prem solutions, hardware, Netiza boxes, whatever our pure data for analytics. We have all that out there and it's a story of hybrid. It's not everyone's jumping off and saying get rid of this. They're saying let's do a part of our computing in the cloud. How does that hybrid scenario work? So there's different ways and it really depends on how you're going to roll out, right? We've seen different ways in which say we have a technology called fluid query. Okay, so this is going to be a situation where I have, and it comes with pure data for analytics. You now are able to, from your logical data warehouse, I can put out a query and that query through fluid query can now go into Hadoop, go into other systems, whether it be on-prem or in the cloud and then basically push that query to where it resides, push where the data resides and compute there and then bring it back. Sort of like a logical data warehouse. Bingo. Okay. And so that's a piece of it there and you're going to see like as we're building out different integrations as well. I mean, I said before, from cloud and to DASH to others, it's certain things work very, very well together, certain things we're building out and we're expanding on. But it's really, when we talk about hybrid computing, there's a whole lot of solutions and different ways to solve that with data that we have it today, whether it be on-prem or... There is, is there a Watson play here? Where does that fit in the data services world? Is there a Watson play? Geez. So there is a Watson play. It's a lot of the stuff that I see today are people playing with Watson services in Bluemix, okay? And different things like language translation, et cetera. And then, you know, how do we bring that in? How do we store that data? I was going to say for right now, more in discussions from the big Watson, because you have to ask yourself, when you say Watson, what do you mean by Watson? Yeah, well, right. I would throw this out there just because there's sometimes Watson versus Watson analytics. Yeah, okay. So Watson analytics, yeah, definitely. So let's talk Watson analytics because that's kind of your group, right? Well, so Watson analytics itself is, we have native integrations with DashDB. You actually can go in there, take their DashDB instance, drop it in there and study it. So we are working across the board and that has... I say it's your group. It's part of the 17, 18 billion dollar, you know, giant analytics business of IBM. Yeah, and that's what's interesting, right? And that's what's fun about this type of role is that there's so much stuff. And you know, I got to be honest with you, I'm learning every day. What about, you know, we've watched Oracle where they've continued on this path where they're optimizing and squeezing every last drop of blood from a turnip to take an on-prem database and extend it with, you know, first engineered appliances and now even at the chip level to get really, you know, top-end performance, obviously at very premium prices. But the cloud story hasn't really come together in terms of, well, in terms of real elasticity or in terms of combining the different types of data management capabilities in a single database. They have that big data appliance that hangs off exadata. When you, now that you have, you know, PL-SQL, the Oracle stored procedure language in the database, when you approach a Oracle customer considering either exadata on-premise or exadata in the Oracle cloud, what's the message and how do they view, you know, a potential migration? Yeah, so I think the first thing is we don't really approach it as a religious war, right? It's not, you know, Oracle versus this or that, whatever. It's, hey, we have something, dash DB or data du jour, whatever it may be in the portfolio and say, what is your business problem, right? If it so happens that, hey, we're hitting the wall, we have issues here, we'd like to do some of this stuff or one of the things would be we'd like to leverage mobile technology or let's say we're doing things on our websites, right? Easy things that are born on the cloud nature. You know, dash DB is a perfect type of store for that along with clouding. I said before that native integration between the two, JSON into data warehouse, all of a sudden it becomes a platform that's a logical fit for pretty much anyone. Now you add into that, hey, we can handle PLC, all that kind of stuff, that makes it a much prettier picture and it's a lot less of the aggressive, you know, hey, you know, come on over for this reason, it's more or less, do you want to do this? We have this here today. If you'd like to come play with this, let's, you know, come try it out. So you created this capability, kind of going to market with it. What should we be watching for as indicators of success, momentum, you know, your objectives for the business? Yeah, I'd say it's a lot more apps based on this. Yeah, a lot of folks, different solutions, right? And this, one of the things interesting between the cloud and building that into cloud services versus let's say DB2 side of it, right? You have a lot of the developer community and you have a lot of the old IT architects, DBAs, okay, they're apparently different beasts, but now they're coming together in the same room, you know, and so part of success is going to be that marriage of the two where we're starting to see a lot more of the on-prem and a lot more traditional players getting more skills and growing into this. Hey, I'd like to roll things out quicker now. I would like to tap into the cloud to bring some of that data in. And on the other hand, you also have these developers that are now saying, wait a second, I want to standardize on this. I want to leverage these data sources. I think the marriage of that is going to be big. I also think is the ecosystem is very important. It's not to say, hey, you must use all of these services and only these services. You use what you need to use. We're going to listen to you and we're going to build on more based on what you want. Would it be fair to say that we've talked mostly about IT and corporate developers? Are there ISV developers of SaaS apps who are looking at this and seeing a cost effective and functionally rich platform to start building apps on? Great question. Yeah, so some of our earlier doctors actually were some folks that are like say intellectual property around, let's say digital apps or studying, let's say predictive analytics around TV showings, movie shows or let's say marketing apps. Yeah, they're tapping into us. And they were some of the first ones to say, hey, because we're creating all this stuff and we're leveraging this in a cloud vehicle. Why don't we maintain all of our services out there? And since we have such a robust portfolio of it, they can tap into whatever they want when they need. And they could say here today, we know that we're going to use X, Y and Z, but we know maybe in a couple of months we're going to want these other services. So they're going to invest on the platform and then as they evolve their product, they can move into other directions as well. But yes, it was definitely one of our first approaches was by that community. What are you seeing Larry in terms of the headwinds that kind of means a lot of tailwinds in this business, right? The momentum, the hype, the marketing and the CEOs reading stuff on the plane saying, hey, we got to do this big data thing. What are some of the challenges that you're seeing with customers and how is IBM helping them through those? Yeah, I'd say one is this is basic block and intact and stuff and you laugh at it a little bit, but it's usability. It's just teaching someone how to do this. I mentioned before the dev advocates at IBM. It's really, you know, talk about, I think about Hadoop four years ago. Hey, I'm in here, I know it's Hadoop. I don't know, you know, a couple of hands are going to be for the data warehousing conference. But people are learning, they're getting there. But it was a mountain that people had to get over. But now all of a sudden you have, you get back to the whole developer market, developer economy, people are bringing their new languages, their new codes and new skills and you're like, geez, where do I start? Where do I end? We're helping people learn. And we're helping, I mentioned before, like these boiler plates and kind of blueprints. We're putting it together. We're taking our experts and saying, let's show you how to do this. I'm like the Lego directions, go from point A to point B to point C. And at the end, you're going to have this kind of a product. And then, you know, from there, you can become the master builder, right? And go into different directions. But I think a lot of it's training and a lot of it is helping people out and helping them, you know, dealing with the cloud specifically. The hardest thing is I don't have a customer, right? Right in front of me. Someone could be hitting the cloud on their own middle of the night around the world. I have to make darn sure that I have the right supply of content to train them and to guide them on what to do or how to do it. I was at the Facebook at scale conference a couple of weeks ago. And a program manager from Google said something interesting because, you know, Amazon, Azure, Google are all building these proprietary services, the big data services by proprietary. I mean, you know, not the Hadoop open source. Right, right. But they're also building integrated development and operations tools that manage the life cycle for simplicity of development and operations of these multi-service applications. Is that something that you're focused on? Where are you in that effort? So sorry, simplifying the- Well the idea that like in the world of 20, 25 years ago we had Microsoft Foundation classes, you know, or Visual Basic to make it easy to write Windows apps. But now we need tools that make it easier to write these sort of big data analytic apps but not only to develop them but to operate them because that's so much harder. And that these vendors were saying, yeah, that's a problem we've got to tackle and it's very hard to do it in a multi-vendor way because, you know, you have to make a whole lot of assumptions. Is that, you know, a part of the vision for what you're building? Yeah, so I'll say part of it is a little bit out of not my direct and my scope but, you know, some of the elements are from a programming perspective, you know, we are supporting and growing and making it easier within a composable service environment, things like Node.js or, you know, Python and other things like that to have the capability within Bluemix that we're following that model to help bring that in and do it from there. Okay, so my last question, which is, maybe give us a richer picture of what those composable services are. The composable, which ones? The composable services in data management and then maybe the things that analyze that data. Oh yeah, so, I mean, when I'm saying composable services, I'm talking about the entire, when I say cloud data services and I'm saying this is our data, like services that are in the cloud, that is the composable service. So things like- Everything on Bluemix. Correct. And so that's the vision of Bluemix in general is composable services. Now my view of that is, hey, we have the data section of it but there's other teams that have their, let's say, NetNew or third party composable services that are in there as well. And that's where I get the whole idea of Lego blocks that this is. Because those are exposable as services as well. Yeah, and they're all, different services do different things. They have different APIs and all that kind of stuff but yes, as a whole, it's this ecosystem of services so I come in and I want to build, I can use this to- And the intent is to deliver that end to end. I mean, that's what your vision is that you put forth. It's like a sandbox for a develop, I mean, I don't say a sandbox but it's a playground for developers to be honest with you. And we're just pumping in more and more stuff as the days go on. With a lot of stuff to play on, yeah. All right, Larry, we've got to leave it there. Thanks very much for coming on theCUBE and telling your story. It's an exciting time for you guys for yourself personally and IBM generally, so thank you. Great. Hey, thanks for having me. All right, keep right there. We'll be back with our next guest. This is theCUBE. We're live from New York City. Pillars 37 at Strata plus a dupe world. This is big data, NYC. We'll be right back.