 Live from Las Vegas, Nevada, it's The Cube at HP Discover 2014, brought to you by HP. Good afternoon from Las Vegas, everybody. This is Dave Vellante, and this is The Cube, SiliconANGLE's production of HP Discover. We're live here in Las Vegas. This is day two for The Cube. We've done a number of discoveries. Very excited about the keynotes that we saw. Martin Fink talking about the future vision, the machine, new modes of computing, HP inventing again. It's actually really exciting, and I'm quite positive on what I've been hearing. Most positive than I've been for a while, so it's a lot of exciting times here. We're going to talk now about HP Software. Brian Weiss is here, and in the center is Randy Carons, both with HP Software. Gentlemen, welcome to The Cube. Good to see you again. Thank you, great to be here. All right, so Randy, let me start with you. Give us the high-level overview. What's going on with HP Software? We've got a new sheriff in town, Rob Young-Johns. We've got him on The Cube before. Passionate guy, great guy. What's going on with HP Software? So what Robert and actually all of us are talking about this week is really how big data changes everything. And really, a lot of focus I think over the years on big data has been on analytics and predictive analytics, but really what we're talking about is how big data has a profound impact on literally how businesses kind of function in every way. How they do marketing, how you drive security, how you manage your legal compliance. And so there's a lot of implications and we're working on kind of next generation software that's going to leverage big data to change those discrete markets. We just finished our big data survey. I'll share some stats with you. Just a couple of very highlight films here. About 300 people responded, only 1%. 1% said big data is a buzzword of unclear meaning. Now, if I asked that question of cloud five years ago, it would have been right off the charts. That was one. The second one was, have you shifted resources from traditional EDW to new modes of computing like ADU? 60% said we have already. Another 30% said we will by the end of the year. So 90% of the base, that is just amazing. It does change everything, doesn't it? It absolutely does. Anything about that surprise you? You know, not really. I mean, what's interesting is if you, if we talk to customers say in our backup business, the data center is under siege. It's getting so much, it's just getting flooded with so much information. Kind of the old models of data protection aren't really working anymore. So we're working on software that can be more predictive in understanding spikes in compute needs and what you need to do to make your backup more self-aware. To really take it from kind of more of a manual process to something that you can automate and almost self-heal in process. So Brian, I want to ask you about an autonomy question. So the world, you go to Silicon Valley, the cloudification, the sassification of everything. So what about idle on demand? Yeah, so this is sort of one of the most exciting things that we've launched in the last couple of years. And David, you're very familiar with sort of what autonomy does to the raw IP is really around understanding meaning inside noisy human information, right? So how do you make sense of a text document? What's it about? What are the concepts in there, right? How do I understand an image, right? Does it contain a face? Does it not? Is it important? So what idle does is it does that kind of near matching, it does a probability modeling, right? Which is more than just, hey, I know that's a match, but it's actually close, it's close enough. So we do this conceptual understanding of human information. And that's sort of the core meat and potatoes of what autonomy does over the last 10, 15 years. Incredible technology. So what we've done in the last, was it really like a year? We've just hardened the daylights out of that platform and we're investing very heavily in what it does and how it does it. And we are at the point now where we are actually releasing idle as a service. Now this is for developers, it's for programmers out here, right? So if you've ever worked with indexes, right? And you've got a project where you say, okay, I wish I could figure out how to recognize a face in this picture. Or how do I analyze the concepts in this block of text? How do I analyze the related concepts to a document that I have? So some of the very special things that idle does are now available as a service layer. So to be clear about what we're doing here, you submit your question and your content, you submit the query, we give you back the answer, okay? So you've got a document, you want to know who the places, the people, the sentiment, the faces in this document, you hand it to us, we give you back an answer. So I can do this in snackable bites, right now. Absolutely. So creating a tech cloud. It's a very different model from the traditional autonomy model, right? Which was so large, complex, on-premise projects, right? We're basically taking that sophisticated technology and opening it up in the cloud to developers to let them go build any kind of app they want to. So last time I talked to somebody, and it might've been Robert about this, I asked him, so I'll ask you again, Brian, what did it take to get there? It's not like you just stuck a data center in front of Idle and put up a credit card machine. What did you have to do to make this a service? It's not trivial, I presume. Yeah, well it's not trivial, but it's part of like, how do we host something where the entire world can hit it and ask questions of it, right? And so managing Idle, how you manage Idle and how you do that at scale is a problem that we had to solve. As well as making it sort of usable from a developer framework, right? So anybody who's a developer that knows how to use it, knows how to get that, to make that easy and simple. So say a mobile app developer wants to do a facial recognition program, right? How do I do that or how do I make that possible? You can develop that in a day. Good guys, bad guys. And I'll give you a great example of this. So in order to get this out there, we've been hosting hackathons. And the hackathon goes like this. You have two days, you get Idle on demand, and what we do as a hosting service is we have some content sources. So we host an index of Wikipedia. And whatever you come up with is cool. And here's the kinds of things people are doing with Idle on demand that we're hosting. The first one we came up with, somebody came up with an application that believe it or not plays Jeopardy. So, in two days. I was gonna ask you what you thought of Watson was saying. Two days, okay, well this is my point. So the core IP of what Idle does and understanding concepts and you know, anybody can use it now. So what they do is they take, Jeopardy has a feed, right? You can get the questions. Take the question off the Jeopardy feed. Submit it over to Idle, wrap some pictures around it and things like that. And 80% of the time Idle gets the answer right. Two days, play Jeopardy, right? Another great one. Somebody did a, we had a hackathon in San Francisco over the weekend. You should challenge IBM to a little. You said it, not me, okay? I'm just saying that the power that this offers to the development community to be able to develop fast, rapid applications that understand human information is we think game changing to the point where we're just making it available free. The business model here is you use it, right? You want to play with Idle? Here it is. The whole strategy right now, what Young Johns is trying to do, starting here up, is attract developers to this platform. So this was a conversation we had with Robert in Barcelona. And he, I think I like about Robert. He's a very straightforward about it. So we, John Furrier and I were, we weren't hammering him, but we were asking him, what about the developers? That is such a key part of any software company. You are absolutely right. It's a gap for us right now. It's a white space that we have to, you know, address it. So what specifically, so you give us some examples here, what else have you done to address the developer world? You want to? Well, I think this is probably the most, Idle on Demand is our most important initiative around this because it takes our, really our core asset, the crown jewels that have taken, they've either been, you have to jump over a cost barrier to get there. Right. And we're saying, look, you can use it. You can use it for free. So we have, it's a true platform now that any developer can use. What you guys at the World Wide Developer Conference? We were. So just to kind of see the trajectory, we really started first talking about it publicly with you and Robert in Barcelona in December. Right. And so since then, we've basically been seeding the market. We've probably staged, you know, two dozen hackathons around the world. And in fact, yeah, last week, we were just at the Apple World Wide Developer Conference. So what was that all about? Just, we have a presence there? Well, we actually, we did a little subversive thing. We did a party in a, in a speakeasy in San Francisco. And, I love the hackers, right? Yeah, exactly. So it was a party like in this, in this basement setting. And we had all kinds of awards and giveaways that would appeal to the developer community. And we had a huge turnout. It was fantastic. We probably had 350 people there, the party, the whole, you know, you have to answer a jeopardy question to get in. Well, no, we made it social, we made it a social media enabled kind of contest. So, you know, they basically had to take selfies of themselves with the idol on demand logo and then answer certain questions about the platform. And then every 30 minutes we would announce a winner over social and then they would come up to the DJ and get the prize and things. So it was, it was fantastic. We had great conversations with loads of developers that night. And a lot of them then came to the hackathon that we hosted over the weekend. And so it's just an ongoing process. So the great thing about talking developers is they want to get stuff done, right? And so the kinds of things that idol on demand will let you do is the common things like text extraction, right? So we don't key view, right? Key view is an asset that we own, we sell it with all of our software. But now, hey, I want you to tell me the main ideas that are in this document. I want you to extract all the text for me. I want you to give me the sentiment of this. That's an API call, and it's free, all right? I want you to tell me the quadrant that the face that's Randy Carons lives in in this photo, give it to me, all right? So facial recognition now becomes an on demand, right? No charge, no setup, no nothing, just use idol. So you're going to come up, go ahead. Oh, I was just going to say, you know, more to your question around what else are we doing for developers. Another very important component of this is working with partners. So as part of the Haven platform, Idol on Demand is part of that overall Haven offering. And so we definitely have done a lot to, you know, educate and kind of evangelize this, these capabilities to our partners. And many of them are now building their own apps off the Idol on Demand API. And you got the HB software has, you know, Vertica specifically has the user conference coming up on August. Will that, will you start to attract developers that or is that more user conference? Oh, we will absolutely attract developers. Okay, so you bring developers into that as well. That's a good opportunity, absolutely, to grow that community. So think about it from a Haven standpoint. I often get asked, well, how do autonomy and Vertica work together and how do I get started? And you heard that we're hosting, you know, Haven as a service. And what this enables us to do is just hit idol and use it, right, with the data, bring it back, do what I want with it as part of a rapid application development. Whereas previously you'd have to figure out how to stand up the servers, load the content, you know, do you want Wikipedia? Someone's got index Wikipedia for you, right? We do that. So eventually what we see happening is people will put more and more of their own data. Like I can feed you lots of my documents and say now I'm going to give you another one to find similar concepts. So it allows you to get right to that idol value very quickly with very low barrier to entry. And the Vertica guys love it because now we can rapidly develop with Vertica. Well, Hadoop, Autonomy and Vertica, to me it makes a lot of sense because people want insights, right? They don't want to just do Hadoop. What's they doing? But they want inside the other piece of the survey was people aren't really getting enough value. A big chunk of R is a big fat middle that aren't getting enough value out of their big data initiatives because they're not being able to extract those insights because they're having to try to invent stuff like autonomy on their own. And we recognize it's going to be an ecosystem of different players who are working together to make that happen. I mean, so like we had MapR and Hortonworks are here this week, you know, having good meetings with both of them. There's a lot of initiatives with Cloudera as well and a lot of the Hadoop community. We know we need to work with them to... Another fit I think, so we had Chris Selendon earlier and he was saying, hey, our philosophy is to keep everything. Don't throw away data. I said, don't say that to the legal CIO at the pharmaceutical company because they don't want to keep working process. That's where you guys come in. You can help me defensively delete data that I don't want to keep. Well, it's also around categorization too. Part of the intelligence is it's something I should or shouldn't keep, right? And for example, one of the hackathons, somebody did a quick application which will analyze a video, right? You take a video document or like a YouTube video, for example, and it will analyze the scenes that it'll OCR. So you want to send me a piece of video content or OCR and give you back the output. And it will then analyze that as to whether it's suitable to watch for your kids. So it's kind of like a real-time video analysis for parental controls. Like is this the kind of thing you should or shouldn't be showing? So I can auto-categorize basically content at the point of creation or use. Right, and fundamentally that's what it does because you're looking at probability modeling and matching, right? And so we can notice that this document, this idea is similar to these other ones so I can go get similar concepts. It can form an index or a mapping of like ideas and concepts. You can't do that manually, right? The only way you can scale is to do. All right, we got to leave it there and get in the hook, but gentlemen, thank you so much. Great saying, but I appreciate you guys coming on. Absolutely. Thank you, David. Yes, absolutely. Take care. All right, keep it right there. We'll be back with our next guest right after this.