 John Furrier, Dave Vellante, live from Hadoop World 2011. Amy's with Nokia. We're going to talk about an interesting case study. Welcome to Hadoop World, New York City. Welcome to theCUBE. Thank you. It's really nice to be over here in theCUBE with you guys. Ah, thank you. So give us the update. Why is Nokia here? Nokia is obviously a huge, huge market leader in phones recent to troubles of late this past year in the news, you know, the memo and, you know, but now, you know, with Microsoft, a lot of things are happening. Oh yeah. Give us the update. Obviously big data, mobile data is huge. Absolutely. Yeah, we have so much going on. So, you know, I think if you think about Nokia, traditionally we've been a product company and we produce data as an exhaust. And now we're really turning it inside out where we create a lot of data, we gather a lot of data, all the sensors that are on the phones and in the world around us allow us to just gather up data and start to analyze the real world around us. So we are at- We can actually do a whole cube event around Nokia's data challenges. I mean, this privacy, you've got geo data, you've got location, presence. Yeah, you hit the nail on the head. So a lot of folks think about it as, from the mobile phone perspective, but we really have two businesses that are coming together. We've got the mobile phone business and we are the world's largest mobile phone manufacturer. And, you know, as you say, we're in a world of challenge. The industry continues to change. We recently had some pretty great announcements. I was in London at our Nokia world a couple weeks ago where we showed some new Windows phones that are coming out in the next quarter or so and can't wait to get my hands on those because they're pretty cool. But also we have a very large location-based business. A lot of it's due to our acquisition of Navtech three years ago. So all of the location properties that you see in Yahoo, Yelp, Forceware, all of the automobiles, BMW, Ford are based upon our maps. And, you know, we view the world as being in the third phase of mobility right now where the first phase was making your landline go mobile in this, you know, with the mobile phone itself. Thank God, finally, killed the mobile. Yeah, absolutely. And the second phase was making your computer go mobile. So we have access to so much when we're on the go these days. And this third phase that, you know, we're having an awful lot of fun with is the convergence of the virtual and the physical world. So there's so much that we can do with data now. We've got augmented reality. Augmented reality, yeah, sharing moments of your life, wherever you are, you know, all those types of things. And that's all about data. It's the classic blurring of offline online, IT, consumerization of IT meets consumer. Yeah, absolutely. I mean, it's really hot. I mean, the location stuff everyone talks about because that's like they think, okay, location, local ad dollars. But it's more than that. You have interaction data from users. You have relationship data. So what are the data challenges you guys have? I mean, you got graph data, you have the API issues with Facebook and other things. So, you know, it's not easy, right? I mean, Twitter is the easiest, right? I mean, that's easy. I think if you went down the list of challenges with respect to data, yeah, we've got them all. And we're picking them off one at a time. Obviously, the biggest one is always privacy. We have traditionally been a company that kind of leans towards the, you know, much higher on the privacy side. A lot of what we do is with Anonymized Data. We're a global company, you know, we sell devices and maps and location properties in 160 countries around the world, which is pretty much every country. And everyone has different regulations around privacy. So that is, you know, the biggest, biggest concern that we have. So what's your agenda here at Hadoop World? Obviously, I mean, big data, but specifically for Nokia, what's on your mind and why are you here and what things do you have going on in the marketplace that tie into that? Yeah, the biggest reason to come here and there's a lot of folks on my team are here as well is to learn from everyone else. You know, we're only a couple years into this big data thing and no one has all the answers. So it's the collective wisdom of the crowd here that's going to help us the most. You know, and we might be coming at it from an angle where we're doing a really good job with respect to privacy. For example, we're, you know, we've heard some talks with folks that are using H-Base in different ways that we're not using it that give us some ideas as to how we might better serve data out and a little bit more of a real-time nature more quickly. So it's learning, it's learning, it's networking. You know, we're no longer in the world of every company is a distinct entity just competing with other folks. It's a total learning thing. I got to tell you a story. So Silicon Angles headquarters actually in the Cloud Air office and I know Amur and Mike and when they moved to Palo Alto is where I live, I popped in and they gave me some free space. They launched this Cloud Air Labs so we were like, oh, you know, because we need you in the video. So I've got to do something with this. So we started an H-Base project. So we run a Hadoop cluster because I'm in there also, I have to do it. I mean, you can't not go there and drink the Kool-Aid. You can't be in that building, right? I'm going to drink the Kool-Aid and do it. And now we're running an H-Base cluster. So we didn't really know what to do at first so that collaboration was the key to success. So do you guys have full on H-Base installed installations? Are you running a big Hadoop system? Yeah, so we have some really, really smart people at work and when things are available in open source, smart people find those things and start to download them. So everyone across the company was faced in their own part of the business with their data challenges. So a lot of different HDFS clusters popped up all over the place. Couple different use cases of H-Base. But what we've decided to do in the journey that we're working on right now is to bring all of the data together and kind of make one data asset for the whole company, manage it centrally. So that's all coming into one centralized cluster at this point in time. It's going to take us a while to get there. But yeah, we're using, you know, every, every pig, oozy, hive, you know, you name it. Everything that's on the list of things being dispatched. I'm renamed them all yesterday. My house is the new one, yeah. We probably spent, I don't know, close to, I had 20 minutes anyway. Just going through them all, it was fantastic. Anybody who wants to get a quick education, go check out the video from Armory yesterday. Amy, I wonder if I could ask you, you talk about the three phases of mobile, you talk about consolidating a lot of the data sources. What organizational implications did that have? Can you talk about changes and how you navigated through all that? Yeah, yeah, so we decided to do one big shift and that's to have a centralized analytics group. But that doesn't mean that all analytics in the company's being done centrally is because we just don't believe that that would work at all. So I describe it a little bit more as a hub and spoke kind of model. So in the centralized group that I run, we do three things. We build out the platform and the whole ecosystem, so at the heart of our platform is Hadoop and the beating heart. And then all around the outside of it, we've got many other tools that we use within the Apache community, but also Oracle, Teradata, Mahoo, SPSS. We got them all, we got them all in that ecosystem. So my team built out the platform. We also work to bring all the data together and centrally manage that data asset, cleanse it, partition it, make it available through Hive. And then we work up the stack and that's where we act a little bit more like almost a professional services organization out in working with the business units to help them do analysis. So that was our big shift, was to create this core that enabled us to keep the data asset in one place and make it available forever. How are you integrating that into the business model? Because obviously, I wrote a post two years ago that data is the new development kid and now I look like a genius for predicting that, but that's truly the case right now. We're hearing that data is a competitive advantage. How are you guys, I mean, you also got to get a handle on it. As you guys look at the business model, how do you integrate that into the businesses, into different products? Is it just you're kind of sprinkling it around everywhere and people accessing it? Yeah, so all the business units now have access to the centralized data and what they're all shifting on doing is, a lot of our products are consumer applications and consumer apps, everything they can do about knowing more about the world and what the consumer's looking for to make a better recommendation to- Treasure chest. Yeah, absolutely. And even just like, AB testing right on the app and all that, all of that's enabled via data. So every single one of our consumer based apps is moving towards a model of being able to consume data that isn't just in their silo, but it's data that they can get from other parts of the business. And then in general, the whole business is moving in that same direction. So you've got the sort of all of the above, data warehouse, big data infrastructure. Did you see the slide that the guy from JPMorgan threw up yesterday? And he sort of said, this is the big source of the day. I thought it was very interesting discussion. And of course the messaging that you hear from Cloudera is no, it's incremental, they were not competing with transactional systems. Do you see that? What do you see there? If you take that, and I think it's a relevant discussion that he brought up yesterday because it is very unclear. What's your telescope say in terms of how this new big data Hadoop is going to affect kind of the existing data warehouses? Is it going to encroach on it? Is it going to be all incremental and just overwhelm it in size? And what's your take on that? So I think two things. First of all, technology keeps changing and I've been in the industry a long time. And if we all don't keep in front of that and understand that everything's going to change and kind of get stuck with one thing and think it's going to solve problems forever, we're going to fall behind and somebody else is going to come out there and compete and win. So that's the first thing. The second thing is I really view it as an ecosystem. I really believe in the open source model, the open interface model and that each one of these components and I look at our, we have a hundred terabyte data warehouse in TerraData and that's a vital part of our ecosystem because we're a 60,000 person company. I can't go around and tell everybody you all have to learn MapReduce now. It's just not going to work. So all of those folks that are using views in TerraData are going to continue to use the views in TerraData. So what we do is we basically now, instead of feeding some of the things that we're getting way too big for that system and are a little too unstructured, we feed them all in through Hadoop, do analysis and aggregation there, push the aggregates out to TerraData so that whole large set of analysts can continue to do their job the way they always did. So that's one example, but I view this whole industry is all about pieces fitting together to solve problems and at different points in time, different vendors and different communities are going to invent something that's going to help and it's going to eclipse something that already exists but you got to move on and go with the changes. We're here with Amy O'Connor, great perspective, seasoned veteran, obviously been in the business, recognizing the new dynamics with Big Data at Nokia. I'd love to follow up with you guys. We're all over Big Data for a couple of years now and I'd love to collaborate with you guys and share information and get you back on theCUBE. The mobile piece is huge and it hasn't really talked about much of this conference but very big, so thanks for coming on theCUBE. We really appreciate it. Let's keep in touch and we want to continue to collaborate. Okay, we're going to have the next.