 Back here live at HP Discover in Frankfurt, Germany. This is siliconangle.com's theCUBE, our flagship program. When we go out to the events and extract the signal from the noise and share that data with you. And I'm excited to be here. I'm John Furrier, the founder of SiliconANGLE. And this is an exciting discussion with autonomy. And Andrew Joyner, who's the general manager of emerging technologies and worldwide marketing for the autonomy portion of within HP. And Brian Weiss, welcome to theCUBE, guys. Thank you, John. Thanks for having us. We have autonomy in theCUBE for the first time. Obviously, I've been blogging publicly about the whole thing, been big proponent of big data. If you follow SiliconANGLE, you know that we love big data. We've been following it since the beginning and we love big data. So you guys, this is your wheelhouse. Obviously, autonomy has pioneered a lot of stuff on the enterprise side, which is very difficult to do. And now with big data exploding with unstructured data and integrating that in with analytics being the killer app of all time right now. You guys now walk into HP now, a good what, year and a half after the acquisition? It has been. You know, what's interesting is this is the first time in HP that all the software assets are underneath one house within HP software. And then within HP software, we have these two really unique software engines. So we have the autonomy engine, which is called IDLE. And we have a Vertica engine. And together, both of those engines were really almost purpose built and designed for big data. You know, if you think about 90% of the information that's inside an enterprise, it's all this human friendly information. It's all the stuff we all consume and love. It's emails and it's blogs and it's tweets and documents. You know, all of that stuff's of a variety. You know, we send it and receive it all day long. So it's of a high velocity. So from day one, we've had all the troubles that many of the organizations had. We've had to designs our systems to handle that. But with Vertica, they're also able to handle the extreme structured stuff. So yesterday for the first time, we talked about to the press and to the community about this bringing together these two worlds, the unstructured, the unstructured. For the first time, we have two unique software assets inside of HP that can handle it. And then what we did is we put in, we had a long list of people from storage, from services to get the alignment. And so it's really been the only HP now. It was a really exciting day for us to talk about we're really geared for big data. Yeah, and obviously, you know, you guys are autonomies in the news relative to the founder and the whole HP thing. We don't want to go there. Obviously I blogged about it and you go to SiliconANGLE and search on some of my blogs if you want it, if you want to get my opinion on that. But when we were at HP Discover in Vegas, we talked with a lot of folks and the buzz around autonomy, yeah, again, this is pre-storm that we're seeing now, but it's a lot of great buzz. And we talked to people from the laser jet division all the way up through the systems guys and services who had just said, yeah, this is like candy for us because now because big data is so sexy at the market level, HP now can integrate some of that magic into some of their infrastructure and systems products. So there's been a lot of excitement. Now rolling now six months forward, where are we today with you guys within HP? Because yesterday we heard that big data is native within a lot of the different products and services group within HP where we're seeing demos, David Scott did a demo on stage and that was awesome. So you see the storage guys doing big data demos, obviously storage and big data kind of fit hand and glove. And then you get the services guys who were going into customer environments where big data is a mandate with analytics driving it. So it's the confluence of real-time information with large-scale on-demand infrastructure. So it's a perfect storm for that paradigm. So good opportunity for HP. So share with us where you guys are internally and just share some anecdotal. I'd love for Brian to give you the color because traditionally autonomy was just a software paradigm. So we would go to people with big data and then the customer had to figure out how to do it on hardware platforms and how to align services and so forth. And what we've done since the acquisition is we made a lot of progress. You're seeing idle and autonomy power a number of different initiatives. Why don't you share some of the ones you've done. So the exciting one you mentioned is the storage integration. And the problem you come up with a big data is that you have to do all the processing outside of the storage layer and you have to trade off the amount of hardware it takes to keep it fresh as opposed to being able to get real-time information. And it's a real tremendous shame that you can't actually use the hardware itself to do the work, the storage hardware. And so what the storage group is on is taking that processing and they're doing it natively at the storage layer. Which when you think about it is sort of a unique HP proposition, right? We can have that bit of the processing for the data that first comes in as it lands keeping it fresh, that work happens on the store all device. And so I look and now move almost 40,000 times faster as far as its ability to create indexes because the work is happening at the storage. Yeah, I mean it was interesting obviously Silk and Angle and Dave runs we keep on our research side. One of the two areas that we're really number one in is big data. We've been covering that space since the beginning. And an area that we just announced some research on first ever report on what we call software led infrastructure which essentially is converged infrastructure modernized. So if you look at the storage as an example with flash changes the equation on converged infrastructure and the software defined trend is really, really relevant. So you mentioned software assets as a standalone that's interesting. But when you bring in software virtualization at storage networking servers, it's a complete game. So obviously that's important. And we just saw yesterday that VMware spun out spring source, cloud foundry and green plum and some other little kind of software areas where there's a automation around social data. So you can see the jockeying for position on some of these guys trying to kind of retool. So you guys are up and running. So what do you think about that trend? Using that example of the VMware spin out you guys are ahead of the game right now. I mean HP no matter how people slice and dice it actually have in place the full triple threat with software defined with Donna Telly's group you got autonomy a year and a half integrating into HP. Are you guys ahead of the competition and how do you talk about that to the marketplace? I think so because I think there's a couple of things that I think you pick up very well. One thing HP is very hard to argue with is that we can do things at scale. Now we ship a PC and a printer every four seconds we ship a server every eight seconds. So we know scale and obviously just the equation you think of a big data is scale. We do have unique assets though. So we're not just collecting and managing the data. We can actually analyze the data uniquely. I mean autonomy gives you a unique answer. It's not a relational database that gives you the same answer at scale. We give you a unique answer and Vertica can from its analytics as well give you something a unique answer. And what's interesting is when you look at information just overall it falls into two buckets. If you collect the information as a source of risk that scares a lot of people. Do I want more information in my enterprise if it's going to increase my risk? We can allow them to get more information because we have the ability to analyze and interrogate that data. We know the policies to apply against that data that you don't increase your risk. And the second thing is obviously it's an opportunity. Nothing aggravates me more than to see companies throwing away data. Throw away data, you're throwing away a lot of your opportunity. I mean any of the companies today if you look at them that are the shooting stars, Facebook, they're an information company, right? We share information, they monetize it. Google, you search for information, they monetize it. The four looking companies are the ones that are monetizing that information. So if you have the analytics and the ability to understand the information you can monetize it. For HP, what's exciting is that's a new conversation we get to have. We've always been known as an IT company. Now we get to go talk to companies about let me help you make money. So Brian, the VMware is an example, is interesting because one of the things that Dave Vellante and I are looking at right now is the stack. A whole new stack is being created. You can kind of see the moves with VMware kind of telegraphing this now, but where is the data layer? So you think about like the middleware model. The old model was compute and apps and it's the middleware. It's the same wine and a new bottle now, right? You got compute, call the conversion structure, software-related infrastructure and apps being big data, mobility, analytics, all that good stuff. But the middle layer is now data. Data portability, all these issues are being re-architected. So I'd like to get your opinion on what you think about that because you mentioned information. One of the main criteria around analytics is real time. You need low latency. And that is inconsistent with the data warehouse and the old business intelligence models. So I want to get your opinion on what do you see happening in that architecture and what do you guys have now that addresses that, the data layer, how that fits in the stack and to the real time aspect of it. Well, there's two parts to this, right? The first one is, as you mentioned, sort of how you do the analytics and where you do the processing and how you handle the compute resources necessary to do that. And the sort of the holy grail of this is to actually put the analytics as close to the storage as possible, right? So if I put it into a storage device, I want the intelligence at that storage device. I want it almost down to something like bare metal per se. But effectively, if I'm storing it and because I'm storing it, I have a view into it that's analytically intelligent. That's where we want to go, right? Now, in order to be able to do that in real time, obviously, there's a challenge about IOPS and you have to be able to deal with the amount. I mean, so what we're going to do in a solid state, obviously I think that'll make a serious impact. And the way in which, for example, Idle works together with the storage stack which we announced yesterday, Store All, is a great example of that. Like a lot of that analytics is happening real time on the way in and so we can speed up the process of keeping the data fresh really by architecting the storage and the analytics at the same time. So let's talk about something that we can relate to which is the Twitter data. So we have a project that was showing a little bit of it where we can get the Twitter data in as such as small low end stuff, better what you guys do, but how do you look at that and how do you store all that massive corpus of information historically and then get all new data? Because one of the paradigms around social data or real time data is storing it and then do data mining takes time, right? So you got to have the ability to, one, do that as well but actually use the new data that's coming in really, really fast and we call that fast data. So I coined the term fast data a couple years ago and that is data that's happening very, very fast, mobile devices or whatever, integrated in with existing data. It's a really hard problem to solve. Can you guys talk about how you address that? Yeah, you know, Twitter is almost ideal for Idle in that we're not keyword based. We're statistically based and one of the challenges of Twitter is how do you find the real conversations about your brand? We did an analysis for McDonald's as an example. The brand team celebrated when they heard their first keyword analysis they said we get a tweet every seven seconds to celebrate. Actually, when you look at the ideas contained in the tweets, 90% of the information was about directions. I'm a mile from McDonald's, meet me at McDonald's, et cetera. The real corpus of the brand was only about 10%. So it's can you naturally understand the information without having to formulate these keywords, figure out who's complaining about cold fries. Once you find that, there's real gold in that, right? Because who's complaining about cold fries can be left to a geolocation. Suddenly you know potentially which franchises are underperforming and you've taken something that was just an information asset out in the cloud to determine which franchises are performing or not underperforming. So let's drill on that. To your point also about speed and volume, right? When you look at an analytics platform, the way that Idle is designed, it's a statistical model, right? Which has an advantage at scale over say language models or other ones. So we can do more volume with less compute power than say a different paradigm for understanding concepts. So we're uniquely positioned to handle that high end. So you're talking about the math in particular. So you can use math to offset raw compute. As opposed to for example a language model where you have to teach it English and you have to say, okay, what's the semantic model? And I got the, the things that follow the verbs are the ones that matter. So we're doing a pure mathematical model which gives us a scalability advantage for stuff. And you're absolutely right. And that stuff gets scaled, right? Well you know, I wish we had more time. I can geek out on this because you know, we love, we design our own stuff. So it's like, we totally understand. But we'll can talk about that later. But the question that's on everyone's mind right now is that in the old data, this is an intelligence market, the ability to do a query was based on a schema. So in the visualization market right now, which is the hottest areas, how do you visualize and get the insights out of the data? It's really, really difficult. So we've been following stuff like Platfora and some companies up in the Hadoop market where you can rapidly ask questions all the time. So rather it's like, it's like the analogy we said, is like imagine doing one Google search a day. That would really not be cool. But most people want to refine their queries as they find out, hey, let's throw away directions. Let's get at the brand as that example. So that ability to do that rapid query, you don't need to do a lot of data science. That's hard to do. So how do you guys address that with the tech? Can you talk about that? Sure. Yeah, absolutely. I mean, there are a couple of things, right? I mean, if you interrogate data long enough, it'll confess, as we like to say. So how do you let the ideas distance themselves? Jeffrey Kasenberg was on stage at DreamWorks. To show you how we get different answers, DreamWorks was doing an analysis about their movie Madagascar. They thought they would see a lot of traditional stuff about families liking the movie or not liking the movie. There was a cluster that stood out and it was children running from the theaters screaming. Not a lot, but there was clusters. When you double clicked, it was Nicholas Cage and Ghost Rider. Why did that show up? Turned out that a movie theater chain had inserted the wrong trailer reel. And it was here as Nicholas Cage with fire on his face on our motorcycle, completely freaked out the kids. So here's the great news. Now they can have an engagement. So they could go to those movie theaters. Typically you produce a movie, you have no engagement. Now they know who is watching. They could reach out to them and help that movie going experience. And it's because the ideas were different that that jumped out for them in using our software. A question on the tech, however. So what's important is that the analytics, the understanding of information and meaning is already done. I mean, this is kind of the point of what Idle does. It understands the information before I ask the question. It's done the clustering. It's done the analysis. And as new data comes in, it's adjusting constantly as to how that happens. So it's more about the data talking to you when you're talking about these types of analytics as opposed to let me ask the question and hope that the compute power can keep up with what it has to do. Yeah, that's actually a good point. And I want to ask the next question on that is, is that you're talking about predictive analytics. And that's an amazing business value. You get the math and the tech behind it. So like the insight around the Madagascar is just kind of a lucky strike. But hey, there's insight there. They kind of bumped into it by accident. But again, that shows that if you use predictive analytics, you get the value. So my question is, Jeff Hammerbacher of Cloudera was on Quora and he quoted, the brightest minds in our business are using big data to serve up banner ads, better banner ads. And that's just a waste. So his point was, big data's not about serving up banner ads or knowing what to do. There's a lot of other values. So final question is, both of you share a perspective of some of the things around big data that are transformative. Is it in the science side? Is it in society? So business value, what big things are happening with big data that people might not know about? I mean, from a tech perspective and some of the innovations that are coming out of it. Glad you started on that. Yeah, you know, first thing I think there's a frontier of noisy information that has tremendous amount of value that we're not mining at all. Video is a great example of this, right? To be able to understand what's happening in video. And that, so those conversations that we're having in that media is becoming something more and more important. And I think, you know, we'll gain efficiencies and insight into that type of information. It's a great example of an unmind sourced fruit. Yeah, and I think what's going to end up happening is we're no longer going to be sitting at our keyboards typing into Google, asking questions and looking at a webpage of answers. I think our mobile devices are going to allow us to find information in the real world. The computers now, the processing power is going to be able to recognize images in the real world and serve up information in a whole new class of ways. And I think that's the frontier. We'll look back at one day, sitting in front of a computer, typing into a little query box. It was a funny way to find information. So you see personalization as a big thing? I mean, personalization is the synthesis of all the information. You think about a smartphone device. It knows your location, the direction, what you're interested in, your profile. There's such a richness of your information that we can now give you personalized experiences. Okay, so to end the segment, I just want to final, final, final question is just quickly summarize where you guys are within HP. How you, how well integrated are you? Share some color around some of the advancements you guys have. You mentioned Vertica and you guys are now bolted together. HP's got a lot of other software that they're developing. How is this kind of fusing together if you can share without, you know, proprietary information just with the audience who might not be aware of how deep autonomy has penetrated the HP fabric? Terrific, I'll give it a penny in a Brian can as well. You know, if you look at Meg Whitman, she's basically made three big bets. Cloud, security, information, and autonomy as a unique software asset is the heartbeat of that information strategy. 90% of what's in an enterprise is this human friendly format. We continue to create and I would never bet against our need and our desire to consume and create more about human friendly information and that's what is the heart of businesses. And so I think what you will see is you will see continued services alignment from HP. You'll see continued storage innovation where they're embedding us with the storage layers and I think you're going to see us prop up in a wider, much broader, whether we used to sell through a direct sales force, you'll see us essentially rise in a number of different forms throughout each day. I'll give you a couple of direct examples like that. So since you mentioned the last discovery, we'll talk about all the various places where we were looking to be embedded. So right now, autonomy idle is integrated with ArcSight. So what we're doing at the security group is looking at all of the log files and when and if something has gone through as an email, idle will analyze that content of that information and tell me whether that event is a problem or not. So when we used to know you sent an email, now we know that that email was potentially about IP, was problematic, so we're returning some of those security things in the daily, a use case for idle inside the security group. Also printers, I don't know if you saw the announcements on Flow, which uses autonomy technology to automatically categorize information as it goes in the scanner and start off a workflow. Yeah, document management is just a killer area. Document management system comes in through the printing. So we're in printing now, we're in security now, we're embedded within the storage group and of course Vertica and Idle is the core software tech. So you guys, so it's safe to say that big data paradigm analytics included and all that good stuff is integrated throughout HP natively. Absolutely, and it's an exciting area for us. I think it's one of those transformational IT trends and I think we're at the start of it. Well, you know, hey, I'm congratulating you, we're fans of big data, obviously all the politics behind the whole deal went down, whatever went down, went down, but still at the end of the day, autonomy had tech, it's being integrated with HP and if you guys can stay ahead of the competition, we'll be watching, so thanks for coming on theCUBE. Brian and Andrew from Autonomy, talking about the tech and how it's integrated in HP and that's the biggest story here at HP Discover is what's going on with autonomy. They got the real deal, it's being integrated in, HP is embracing it, so the chins are up and they're not afraid to talk about it and we're happy to have them on theCUBE. We'll be right back with our next guests. Thank you. Thanks so much. Ecology and delivery undergoes a tectonic shift. Examples of these shifts are all around us today. Mobility, so.