 Live from Boston, Massachusetts, it's theCUBE at the HP Vertica Big Data Conference 2014. Brought to you by HP with your hosts, John Furrier and Dave Vellante. Okay, welcome back and we're here live in Boston, Massachusetts. This is theCUBE, our flagship program throughout the event. I'm John Furrier, my co-host. Dave Vellante here, Jeff Vist, the VP of Solutions Marketing at HP Autonomy. Welcome to theCUBE. Great to see you. Give us the update on big data, meeting business convergence. We were talking on the keynote earlier, it's really on theCUBE. IT deployment of big data and then putting it to reality. A little bit of disconnect. Business managers want results faster. IT thinks they're checking the boxes. You guys are playing on both sides. You straddle that with autonomy. You're up and down the stack. What's the update? Well, there's a couple irons in the fire, per se, to use a metaphor. A big, big push that I think is worth talking about today is idle on demand. That is gonna take this genre into a whole new area. But in addition to that, we're seeing the emergence, and I think you start to see this one marked, start to mature and grab hold of purpose-built applications. One big bet, and we're showcasing it here, is a healthcare analytics big data platform, specifically for looking at clinical data. Not as a turnkey full solution, but as a platform that others can innovate around. And that shows you the spectrum from something that is far more accessible, that developers that would never touch a technology such as idle in the past can access, and then something that you can take directly to line of business managers, in this case, in healthcare, and have them be productive in delivering results in weeks. So on the analytics side, I got to ask, because this is something that we've been talking about. We've been talking to Jeff Kelly, who's did this big study. And years ago, Mike Olson was the CEO out of Clutter at the time, was like, oh, the application economy's coming. We just never saw that happen coming. There was never a tsunami of big data apps. It simply was the killer app of analytics. Certainly visualization and analytics seemed to be the primary game in town, in terms of value. What do you see coming down the pike? Obviously, you see that as well, or not, share your opinion on that. And then when are the apps coming in, or is just big data gonna be native in all this? Well, yes, and yes, I guess, okay? So a good political answer. I've been in analytics a long time, and there's a lot of inertia, and inertia doesn't go away quickly, by any means. So there are IT organizations that have made substantial investments, data warehouses in the tool sets, and I've seen again and again where you bring in an app, maybe a proof of concept, you show it to that line of business manager, their eyes pop out, the value is immediate, and you think, this is gonna be smooth sailing. We can just mail it in at this point. And then what happens is you show it to IT, to the practitioners, and the first question isn't what value do we get it, how quickly can we scale it, it's what technology do you use, and how can I build it? And that in a nutshell captures I think the roots of where we come from. Now what is gonna change that from doing it kind of grow your own? Well you need IT because you need the data, and the data comes from those resources within ID. That's what big data is about, in fact going to more diverse resources. So it's not about even if you move to the cloud cutting out IT, it's really about embracing it. But for me the change that we're seeing is taking solutions that are far, far easier to weave into your apps, and making it intrinsic to the apps themselves. And that's where it was a yes, yes answer, is that things like search, nobody's running around saying number one priority is enterprise search, but if you can't find the right data and the right relevant information, what's the point of doing any of these big data apps? It's now though that those kind of capabilities, not just search, but other functionalities need to be intrinsic and accessible. I mean search is a great example, we were talking earlier with Etsy and also with Wayfair and e-commerce companies, they're seeing the search, they analyze the search like it's nobody's business, because that's key and important part of that thing. But it's not coming from Google anymore, it's coming from apps and all these different omnidirectional points out on the web and the internet. So that's certainly a big shift, right? So whether it's enterprise search, there's a new vectors of discovery, that's a huge thing. Now it seems like that's a big data problem, but it's not a big data app, or is it? This is the weird thing, it's like what is big data? Is it a category or is it just native and everything? Well search is going to be fundamental to every enterprise app and we all know that most search stinks, right? And you're being kind and it's early in the morning. Right, most searches is absolutely horrendous, it's useless essentially. And what a lot of people don't see is, and especially consumers in line of business people is they go to Google, they search on the web, they use Siri and they get a great result. How many times do you go past the first page of a Google result? But when you're dealing with enterprise data, where you didn't have five million people searching before you and searching for one thing, a link, a hyperlink that is being refined again and again. When you're going to enterprise data, you have to look at the payload, you're usually not looking for a link, and you're right, you go to that proverbial white box and you're usually wasting your time and that is why we're doing Idol on Demand because this is about contextual search. You see a lot of terms thrown around, but it comes down to the simple idea that I have to interrogate the data, I have to understand relevancy, there's a distance between ideas of unstructured information, and I have to have the right technologies that's going to find that. Very different than running a popularity contest, which Google does very well, but that's all they do. And you suggest, and I'm inferring from what you said, Jeff, that the Idol on Demand gives you scale and gives you data and insights that you can then apply to improve people's enterprise search capabilities, is that right? Well, I think it's a game changer and I'll tell you why. We've had Idol for over 10 years and it powers some of the most robust applications, e-discovery, compliance archiving, heavy duty stuff, but it's a heavy duty app. You need a lot of hardware, you need to do what we call build the index. You have to analyze the data, usually in batch mode. Once you have that index and we manage in place, so we're not shifting the data around, which is another key part. I think a number of the actors, John, are they're moving the data to analyze it. So they're saying as long as you put your data in our little piggy bank, our data repository, all's good. That's a deal breaker, because you're going to move the data, forget it. For big data, the size of the data. Why not move the compute to the data or move something else around? Or how about if you can link and point to where the data is so that the data can be diverse? But, and we've managed in place for a long time, so that's not an issue for us, but I tell you what was the issue. The computational intensity of building these engines into your app, that, unless you have a really, really thorny problem, wasn't doable until now. Now, Idol on demand is where we took those functions in Idol and created APIs. Cloud-based APIs, a simple call, today it's stateless, we'll move to stateful. There's 26 APIs ranging everything from a basic search to face detection, to finding similar documents based on context and not keyword. You could not get those until this year unless you put that engine into your application. Now you can make a call to a private or cloud-based application with Idol on demand. And now that means I can take what I could, would have had to put literally thousands of man-hours into developing, now it can be an API call through a web app. We were just having a very inspiring conversation prior to get here. Jeff with Alan Nance from Philips, huge company, obviously they spend a lot of dough on services. And of course, you guys are a big supporter of those guys, but the issue he has is, clearly I asked him all this transformation stuff, I said, what's the outcome of all this? All the stars line up, and he pulled the string and everything happens. What's the objective? What's the outcome? He said, I just want to get close to the business. And that's the big data challenge with the analytics, right? It's really, is that, do you, well, and it's not obviously cliche to say that, everyone wants to get close to their business, but how do you make that happen? What's your experience? Well, you don't, there's two sides of it I think. One, and you see a lot of focus on this is doing what you already understand faster. To be able to do it in near real time or real time in context of your business. And that may mean one of your online users clicks and I determine in a millisecond period of time what kind of adder information to put up. In other cases, it's not gonna be milliseconds. It's gonna be many seconds or maybe even minutes may be suitable for giving that information. So time has context there. But the other side, and I'm not trying to wax philosophical here, is you don't know what you don't know. And the trick here is to let the data tell you and reveal to you the insight. To allow you to see breaking trends. Often we refer to this as looking for patterns of information through weak signals. Now that is a very, very powerful concept. And if you want to get closer to your business, I would argue, yes, sometimes it can be reporting in the general sense. Sometimes it can be clicking through your supply chain and trying to understand and have transparency all the way through and understanding in minutes what may have taken several weeks before. That's part of it. But I would argue that is balanced out or maybe even overwhelmed by the ability to look at weak signals in your data to understand your customers, understand how you're operating and act on that before your competition does. That to me is the game changer. That's what big data is really all about. And that's something you can only do if you have the right technology and you have the data sets to be able to bring that image together. Almost always it does not come from one siloed index. One data set. It's usually diverse data spread around that you cannot connect the dots any other way through big data. And then you can create a virtual index essentially, right, is what you're doing. Is that correct? Well, yeah, you can, the whole idea is to look for relevant and irrelevant information that comes in different structures, different formats and then be able to contextually understand what that information is discussing and talking about and then make the join between them. Not that you're merging it together in one big MDM, but you're starting to connect together something as simple as a Twitter feed and your customer support emails, right? I can't tell you the number of companies I talk to, they have social media analytics. There's lots of great tools out there, we make some. And then they have usually some sort of response, multi-channel where their customers are screaming at them, praising them, reviewing them. It is a very, very small minority today. I would say less than 5% that can join those two data sets together about how you're being discussed on social media, which is not in your data center, right? And then the information that's coming in. It may sound incredibly obvious, but today that big data problem is sitting right in front of virtually every company and a far majority capitalize on it. And that is a great example of what we need to do. How do you tell people to jump on that opportunity? Because let's drill down on that. That opportunity is sitting there. What's your advice to your friends and your customers about that opportunity? What do they got to do right now? Like seize the opportunity, critical, top three things. Well, number one, I think they have to break their organizational silos because I can guarantee you, not guarantee you, but I would say in most cases, those data sets reflect the organizational design they have. They probably have one group of social media analytics, probably the average age there is 16 because they understand that technology and then there's probably another more ERP classic traditional data set behind those glass walls, running the financials, the heavy duty mission critical systems. You got to mix it up. And so number one is mash up your organization so that you are not siling your data like you are your organizationally. And then I would say number two, you have to start finding that right platform. It may not be one technology, but you have to find where the joins are. I think the breakthroughs in innovation are usually at the edge. So you have to look to where the edges are and then start to think about how can I unleash the data. When I start to see big data initiatives that are consolidation, land grabs if you will, let's put all that data in one place, make it big and count the terabytes. Usually you see organizations are going to be less successful. And the whole concept is antithetical to big data, which is like John said, ship the function, not the data. So how does they do things like mobile affect the ability to find patterns through weak signals? Well, that mobile is the great kind of equalizer that allows people not to be sloppy. Because number one, you do not have a 28 inch or 45 inch screen that is sitting there that allows you to kind of tool around that data. All of a sudden you have to be able to do something in a really elegant, intelligent way. So you have to push the intelligence upstream so that you can work through your device. I'll give you an example. With Idle on Demand, well two quick examples if I can. One incredibly fun, probably has no real value in the world, but I gotta share it. And the other one probably useful. The useful one is one of our hackathons, one of the developers created a Superman search and how it worked is they held up their mobile phone, made an S at an article. They typed nothing, nothing. We could do a contextual search in that article and then find similar information that we could bring back. No keywords, no metadata, and all you needed was a finger, a digit, a stick, and you could do a contextual search, okay? That shows the power and the necessity of mobile where maybe if I had a big screen, I could be clicking around, I could be bouncing around, I could be lazier, I could be sloppier. So that's the good example. The one that's fun and probably doesn't have any real value in the world, but blew us away at the potential of bringing this into mobile footprints is one of the hackathon teams had a Nerf gun. The infamous Nerf gun that shoots foam darts, this one was electronically operated. They slept a Raspberry Pi board on it with a camera. The Raspberry Pi had a Wi-Fi interface so they could make an API call to face detection. They pointed the camera at several people and triggered them or tagged them rather as good or bad people. And then they walked around the room after doing this with this Nerf gun. It would look at you and then automatically determine whether to shoot you or not. And it would, it scared the heck out of people. They did this in one day. They did it in what you could argue a mobile footprint, not a phone, but in this case, a piece of weaponry that was obviously a toy, to show you how they could bring face detection, image analysis into a footprint that, while in one sense was funny and in another sense gave you a sense of what you could bring to a footprint that you couldn't do before. So I think we're at any one out one of a long game that's about to come where we're gonna see applications enhanced and powered by big data capabilities but in ways that we generally didn't think about. This is far, far different than running your ERP report to see your supply chain optimization in 30 seconds instead of 30 minutes. That has value, but what I'm truly excited about is this next genre of capabilities. And that's what Idle On Demands all about. That's our big bet on big data. I don't, I think it's more than lots of data faster. I think it's really about a new generation of apps. So I got to ask the cultural question because you've been in the business for a while and David and I were riffing on this in stages when we introduced this concept of born in big data. Really kind of just pivoting on the mobile first. Microsoft's got this thing called cloud first, not that they invented that, but that's their drum beat right now. So we have the notion of data first. So this born in the cloud, born in big data mindset, the kids, the young guns or the new folks who are coming in without the baggage of IT, clean sheet of paper, they're looking at no barriers. They're born in big data. Is that the future? Is that the requirement? You're seeing, I mean you can see this, born in big data and not born in big data. And if every tool for the job is, there's a tool for every job, well if you're a hammer everything looks like a nail. So you don't go in with this, right? So what's your take on this? So how do you balance that new generation of users? Well the new generation just comes to a hackathon, okay? We need to get a lot more kind of silver-haired people at hackathons. What is the born in big data? Describe it in your opinion. What is the culture like? Well, they think differently. They think differently and they are far more risk taking in terms of looking at data sets and not feeling like they can screw it up. And they're far more judgmental about the value of the data in ways that I think the previous generation view different data differently. I think the previous generation thought, the closer you are to the data center, the more valuable, the more trusted. I think this generation thinks opposite. I think when they look at the cloud, when they look at social media, when they look at the interactions that are there, that are unmanaged, they actually put more credence in that. And the closer you get to the data center, and this is a little extreme, I'm painting a black and white, I think the less credible, the less timely they think that data is. So it is a reverse mentality of outside in versus inside out. And that means where they start with big data starts at a different viewpoint and they are polar opposite. And if I had to capture a metaphor or- So app down to infrastructure, so maybe that's app view. Well, that's the second thing is they don't really think data app, okay? I think the previous generation thought, I build the app, it maybe consumes some data and then if it's a good day and the sun's shining, I'll do some analytics, I'll deliver some insight. And usually it's the form of reporting to just validate what you already knew. They don't think in those kind of parsed like terms, I think they think about them in the moment, what experience can I deliver and then they build out from that standpoint. Now, naturally that plays in the consumer. I think the big question is as Siri invades the enterprise as these tools move in, and I'm not just talking about nice user interfaces, but I'm talking about allowing anybody in that organization, obviously with access rights, which is key, to be able to get the insight, to be able to find the truth. That is very threatening, not just to IT, but probably to every CEO down. There's knowledge, there's power. And I think that is the way that's coming with the new upstarts. Yeah, let me see that result before you send it out kind of thing. Where are we, Jeff, with idle on demand? You mentioned 26 APIs. Where are we with customer uptake? What kind of metrics can you give us? Well, if I can say so, this is HP being not HP, okay? If you go to idleondemand.com and what I mean by that is we are full open transparent about this. We put it in early access. It is available with no monetary commitment. You just log in and join the community. We have over 3,000 developers in the community growing at about 30% per month. So huge growth and huge interest. And it's not that we have the only APIs in town. I don't wanna claim that, but I would say we have the most robust, diverse API set that you can find. You'll find one guy that does enterprise search, one guy that does face detection, one guy does this. You go and see these 26 APIs which are gonna grow to over 50 by year end. And introducing pipelining, introduce advanced capabilities so it's more than just an API call. The feedback we're getting from the developers is we haven't seen this treasure chest of capabilities in one place ever. And so what you're gonna see to answer your question is top coder challenges. We're pumping a lot of money. If you're gonna buy stock, buy stock and top coder because that place is just a mecca. It's a petri dish for developers to come in and engage and man does it make big companies like us honest. So there's top coder initiatives going on. There's hackathoners going on. We have the idle innovators contest going on. We've opened up the kimono to engage not with practitioners that are obviously critical to HP's success and HP's business as an infrastructure company, but in this case we're talking directly to hardcore developers. And you go to idle on demand, you play with it, you give us feedback. And we are doing innovation really from the outside end. I've never seen it in HP. Every analyst I talked to said they have not seen this kind of approach in HP. It's always been grow it in HP labs, have a lot of PhDs and then show it to the world. This is the other way around. You know Robert Young-Johns we've interviewed a bunch of times. We love having him on theCUBE. He's one great guy. He's very talented, great content. But one thing we notice with what he's doing, what autonomy is doing in the Vertica is you guys are aggressive. I mean you're like doing the HP thing, saluting the HP flag, I get that. Things going on there, but you're in on the ground making things happen. You got Chris Sellen saying, hey don't ask us, ask your peers. And this is the new way to do business. I mean and this is, you're not only eating your own dog food in the sense of, you know, these expression with the product, you're actually walking the talk on the culture. So, you know, I think you guys are doing a great job. I think you got to get the word out on the visualization side and the interfaces are everything. I agree with you. This API notification world is going crazy. So, you know, the interface is everything. You see B from Etsy said that on theCUBE and user experiences are changing at the developer level, at the IT level and the whole new analytics. So, with that in mind, I'll give you the final word. What is the goal for the next 12 to 24 months from a product, market perspective? Not so much performance, but like just objective, on board more developers or what else are you guys working on? Well, number one is to push the envelope, okay. We have incredible technologies. We have very little to lose by taking this technology, expanding it. And if you want to have a platform and HP at the end of the day is a platform company. And I say that proudly, proudly. But the way we deliver a platform so it can be consumed and innovative, that has to change. And that's what Idle on demand about. That's what this Vertica conference is about. This is about engineers talking to engineers. And that is HP kind of coming full circle. You know, the HP of the old, you know what a sales call was like? It was like, bring in the computer and don't take out a PowerPoint, open the cover. It was an engineering company. I worked there for nine years and you got the manual, here's the parts list. You know. And that was, you know. Hey, which part never you want to look at? You're using those resistors? No, they were geeking out. They would geek out. Sales process. Yeah. And that, you know, whether it's Meg talking about reinvesting and innovation or us breaking maybe the traditional rules of a nice, safe HP where we're mixing it up and mashing it up. That I think reflects not just what we want to do, what we need to do as a company. And- Which is known for reliability, trust. They had that engineering geek culture before it was cool to be geek with Dave Packard and Bill Hewlett. Great to see them mentioning the keynote. But, you know, also they were humble. HP was kind of a humble company. Well, you know, when you're humble you can be humble for a couple reasons, right? When you're not humble, it's either because you have the best mousetrap in the world and just everybody else is going to find that out. And if you can pull that off all the credit to you but there's a lot of people that are not humble if I can be polite, that don't have the goods, okay? They have- That's why theCUBE exists. We want to highlight that. That's what we do. Well, there you go. Well, the customer's highlighted. We had the Phillips here basically called out, laid out, who had what? And, you know. There's nowhere to hide. There's nowhere to hide anymore. So when I talk about HP being aggressive it's also we want to have the goods. HP has always been a company that pride itself that we deliver real stuff and we take care of our customers. And if you can do those two things, okay? And I don't know if that's a bumper sticker, you know, real stuff and customer success. But if you deliver on those two things and you're straightforward about it then I think a lot of people can get a lot of value out of you. And that's what we're trying to do. It's just now we've turned on the turbo boosters, opened up the doors, and you're seeing an HP that is just beginning and we're having a lot of fun doing it. And that's the big focus right now. Jeff, thanks for joining us on theCUBE. We appreciate it. We got to go. We're getting hooked on time, getting the hook. We are here live in Boston, Massachusetts. This is theCUBE. We'll be right back with our next guest after this short break. Thank you.