 Hi everybody, we're back, this is Dave Vellante with Jeff Kelly of Wikibon.org and this is the CUBE SiliconANGLE's production of the Tableau Customer Conference 2013. We're here in the nation's capital at the Gaylord Hotel and we've been going wall to wall for the last two days talking to customers, talking to executives of Tableau, independent consultants and observers like Ray Wang. Today we're talking to several partners as well. Rick Tam Daniels is here. He is the vice president of technology at a company called Attivio, a Boston-based, Newton, Massachusetts-based company that is doing a lot with mixed content, big content, bringing together transactional data and other types of unstructured data into a unified view partners with Tableau. Rick, welcome to the CUBE. Great, thanks for having me, Dave. Yeah, so let's talk a little bit about what you guys are doing here at the event and what the relationship is with Tableau. Yeah, so we've had a long-standing relationship with Tableau, this is actually my third TCC and Mass is one of the most exciting events that I get to go to every year because this crowd is so enthusiastic about the technology. It's unbelievable, isn't it? Yeah, it's unreal. But yeah, so really what we're doing in the ecosystem here with Tableau is we're delivering insight from the world of human-created information. So think about emails, documents, CRM case notes, survey comments, there's a lot of valuable business insight in those sources that's highly relevant to a lot of the analysis that's already being done today with Tableau and we're able to bring that information in, Tableau connects to us like any other data source and we're able to create rich visualizations to help businesses make better decisions using all relevant information. So you guys got a booth here. Yep, very exhibitor. Talking to the customers flying by, what kinds of things are they asking you? I mean, I know, I mean, I was asking some of the execs yesterday about how do you deal with unstructured data, how text analytics and the like and they said, well, many of our partners can help us with. Is that, for instance, a problem that you solved? Do you get that question a lot? What kinds of questions are you getting at the booth? Yeah, absolutely. One of the things we talk a lot about just kind of in a little bit of a way of background is the concept of what's being called big content now and Gartner's out there talking about this is an idea of, there really are different technologies and techniques to get to the different parts of the big data ecosystem and kind of lumping them all under unstructured or big data really doesn't give you the nuances and the technology needs to effectively get information. But, and so there's a lot of talk about, you certainly questions about social media. Big content does that though, right? Yes, absolutely, of course. Okay, explain what we care about though. Yeah, so there's a lot of folks out there on social media, social analytics, but the reality is for a lot of businesses the text-based information that's the most relevant is sitting within systems behind the firewall. So we got examples, we're doing a demo at our booth that folks can check out where we're looking at one of the billing problems in healthcare. So hospitals, the way they get reimbursed is by properly tagging patients with the diseases they have to get funds allocated and but they're always mistakes. There's always issues in the business process, clerical errors, but where's the point where you know for sure what a patient has? It's when they're interacting with the doctor. It's in the doctor's notes. So we're actually able to analyze the doctor's notes, identify what diseases and conditions they likely have, and to have a level that's to visualize what is our exposure across the entire patient population? What, how much revenue are we missing out on because we're not coding people correctly? So you're saying you can read the doctor's notes, can you read hieroglyphics too? Fortunately, they're either dictated or typed in today. Jeff, thank heavens for the EMR and real use initiatives that are going on. Okay, so these are actually doctor's notes that are in electronic form. Okay, you haven't solved the speed of light problem. We have not solved the doctor's hand writing problem. That'll be a new company, I think. Okay, so Rick, you said that the real value is what's behind the firewall, and I wouldn't dispute that for the vast majority of companies, that's true, but you're sort of behind the firewall or not agnostic. Exactly. You don't care, right? If I'm a customer and I say, well, I really care about social data, you can help me with social data, right? Absolutely, yep, and we'll get, hope you get that social information in, but we'll also say, okay, if you're trying to do customer experience analytics and understand what are our customers telling us by looking at the social channel, for many industries, that's one piece of the pie. Email is still a predominant communication mechanism for customers interacting with businesses or even call center notes. All that stuff has a broad coverage across the customer base that complements what's happening in the social ecosystem. So the challenge I think a lot of people have with the Tivio is you're a platform and you can do so many different things. We've talked about that before on theCUBE. It's so powerful and we can do this, we can do this, we can do this, we can do this. Where are you finding the best traction? I mean, you guys just did another big raise a little while ago, you got some great technology, you got a nice customer list now, so you've taken your time to build up pretty strong companies, so you're getting footholds, I would imagine. Can you talk about some of those areas where you're getting traction? Yeah, on a high level generally it's in the business value side of the big data conversation to be honest. It's going in and saying to business users what information is relevant to how you conduct your business day to day? Let's help you analyze that. How do you make decisions today? How can we bring in the emails to help you better understand what customers are excited about or not excited about or unhappy about? So in terms of different industries, financial services has always been a strong industry for us because there's such an impact when you move the needle a little bit by bringing more relevant information. Even a tiny amount can be a huge benefit. We've also seen a lot of manufacturing, so warranty claims analysis is popping up all over the place. Looking at CRM cases, supply chain comments, in telco it might be technician notes out in the field to understand why people are having cost overrun servicing a cell network. All these places where you can have a real impact on actual business problems, the core business KPIs, are bringing in more relevant information, more valuable context to the analytical process, the decision making process. And talk about what makes a TIVIO unique. I mean, obviously, well, from my standpoint, it's the ability to deal with the different textures of data, that's unique, but maybe in your words you can describe what's your big differentiator. They're kind of two major pieces to it. One is getting deep insight from the world of human-created information, be able to go and get it where it lives, understand it, analyze it, and make it available to the BI ecosystem. So that's an analytics component. There is an analytics, text analytics, natural language processing, and then on top of that, we've created a unique data store that lets you in a very agile way connect the insights from that world with insights. Let's say you're getting from Hadoop with the structure data that's relevant to business. So you get that true 360 connection without having to do a lot of data modeling, a lot of data modeling overhead. Okay, so that was one piece of it, the ability to deal with things like analytics. There was another one that I- Yeah, and basically it's to connect the two. So it's to take the structure data, connect it with these unstructured content, unstructured data insights. So on number two, what role does visualization play in terms of painting that picture? Well, I mean, it basically puts a face on it, right? And Tableau is fantastic at showing, you know, really when visualization is about showing you what you should care about, right? It's about what's the outlier? What's the trend? And then you want to say, well, why do I see this trend? Why is the outlier there? And that's an oftentimes the unstructured content, the big content provides that critical context. And if you can visualize that as well through common key concepts and sentiment, you can get a complete picture in a single dashboard. So what kind of integration have you done with Tableau? So it's actually very straightforward. We're unique in the market in that we support ODBC and JDBC and SQL access to very much a store that is, I guess we've fallen to kind of the no SQL store bucket. But in addition to SQL, we support a full text search language. And so we're pretty unique in that Tableau when they look at information we've collected, whether it's emails or documents, it's tables, it's rows, it's columns, that Tableau users are already used to analyzing. They know how to use the technology. It's very easy to pick up and adopt. So another thing about Tableau, again, it's this platform, it does a lot of different things. I would imagine you've got people in your ecosystem or your customer base, they might use bits and pieces of that platform. You mentioned search, you mentioned text analytics. So are you finding that? Are you finding that people are, yeah, hey, I want to use Lucene search, or will they use your search because it's more powerful? What are you seeing in terms of the components? Are people buying the whole enchilada or are they utilizing piece parts of your technology? I think it's certainly, most of the platform is really what people are interested in because even if they come from the search side, they usually have found too many limitations in the search world, in the search technology base, and they realize that because we can look at data from a relational standpoint, we can model data with the kind of same mapping, that the cardinality of a database, not to get too technical, but the idea being that they're saying, you know what, not only do you want to search, we want to relate, we want to analyze, we want to visualize, and that's really why we were founded in the first place is that there were these use cases coming out of the search world, out of the BI world where we wanted to cross over between the structured and the unstructured domains, and that's what our technology's been designed to do. Rick, so we're here just outside the nation's capitol, so I'd love to talk a little bit about what are you seeing in terms of adoption trends of both your technology tableau, but even more generally just kind of the, as the big data world kind of evolves and we're starting to see more real applications being deployed. Is the government sector a place where you guys play and are you seeing anything interesting there? Obviously it's in the news a lot for the intelligence flap at the NSA and other things, but what role do you guys play there? Do you have a lot of government customers and are you seeing pickup in that area? Yeah, government's actually a great use case as well. I mean, obviously intelligence, the high profile stuff, but government in a lot of ways acts like a business, so they have some of the same systems, support tickets and things that you want to analyze as well, so it's applicable all over the place in the government sector. And we have done some work with some, from various entities around the world and that sort of thing in the intelligence space, but that's probably all I'm supposed to say about it, so. Sure, yeah. I understood, but yeah, I mean, good point. I mean, they have a lot of the same issues. You know, in business they talk about the 360 degree view of the customer, and in government, for good or for bad, the 360 degree view of the citizen. What about other industries or verticals you're seeing? As this kind of this conversation around big data evolves, are you seeing any particular pickup, you mentioned financial services, are you seeing any more of what you might call more traditional or maybe not as tech-savvy industries, what the manufacturing, even agriculture, starting to pick up in the use of this kind of technology? Yeah, manufacturing, absolutely. We're seeing lots of activity there, healthcare too. Healthcare is huge, especially with the Affordable Care Act, there's lots of mandates now to go to kind of outcomes-based medicine, outcome-based care, and a lot of that's going to be measured by textual interactions, patient feedback and surveys, and it needs to be captured and analyzed for reimbursement and how the whole economics of healthcare may be based around this type of information. Right, absolutely. I think in healthcare there's huge opportunity to improve outcomes, but there's also, based on due to regulations, the requirements, reporting requirements change, how does Ativio help healthcare organizations actually navigate that all those regulations and those concerns around privacy and security related to HIPAA and other regulations? Yeah, that's actually one of the great strengths of our platform is that we've a really, really elegant and mature security model. So controlling row-level-based access to information is something that we do extremely well in very flexible, very flexible, very complex security infrastructures. So you have oftentimes not just one simple authentication authority, you have a whole bunch you have to deal with, and our engine, because of its agility, lets you handle that native security model and apply it in a very performant way against the data as users are querying and they're interacting and requesting information. Let's talk about the company a little bit. So, you know, we're obviously here at the Tableau event. We talked a little bit yesterday about the, you and I off-camera, about the Ativio's basically partnering strategy and how you're looking to go to market with some partners. Tell us a little bit about your strategy there around, obviously you've got a relationship with Tableau. What's really your partnership or alliance strategy? So our strategy is to create our view of what the big data space should look like and what a good portfolio of technologies are to deliver on all the pieces of relevant information out there and bring them together to the end user for a business purpose. So when we look at, you know, we also partner with Informatica, for example, the stuff they do on the structured data side of the world, those are hard problems that have been there a long time, they're not going away. So their technology is critical to be able to get a clean view of that side of the world and then we can connect that up with the new view, the new information, the big content side and the analytics that we're doing to create this enriched view, the highly contextual view that is really starting to move the needle for a lot of companies and a lot of different industries. And so what about on the database side? I mean, you have some functions that seem to put you in a little bit of competition with some of the database players, but also I can see them being complimentary in a lot of ways as well. Yeah, absolutely. What's your kind of relationship with that growing ecosystem, not just the relational world, but certainly the NoSQL world and the Hadoop world? Yeah, so we partnered with Cloudera in terms of the Hadoop vendor we work with, but you know, we're a Java based platform, so it plays nicely with all the different Hadoop technologies that are out there. We've also worked with some of the analytic database vendors. Our text analytics live in a pretty unique part of our infrastructure. It let us, in addition to bring the information together with structure data, we can actually push it to other places as well. So if someone has an analytic database and they want new text-based dimensions, like text analytics, key concept sentiment, we can absolutely deliver that to those environments as well. How was social data playing into this? I know you were saying a lot of it's sort of behind the firewall, but there's plenty of social data, social interaction going on behind the firewall. We were asking Nate Silva about this notion of crowd-spotting. Essentially the idea being that you've got a lot of interactions in the crowd, the wisdom of the crowd, et cetera. But are you able to, are you finding that customers are able to actually get value out of that data? Do predictive analytics out of that data? I wonder if you're seeing that trend start to emerge. We see it in specific industries. So I saw a great survey recently, I think it was about six months ago, looking at what is the social media adoption on a customer basis? So what percentage of customers interact with companies and spy industry through social media? And outside of retail and entertainment, that's probably about 40, 50% of the customer basis. That's kind of a pretty high number as well. But when you go to financial services, it's 10%. I suspect a lot of that's retail banking in particular. So for our institutional investor customers, that's not as relevant to them, but they do want to tap into customer conversation, which brings us back to those internal sources. So we really think about, holistically, it's the internal plus the external. You obviously need social, it's important in certain industries, and you don't want to ignore it. There's a lot of value there, but it needs to be brought together with more contextual information and provide that true analytical that you can use to make decisions. Okay, so it's taking that one step further because Nate Silva was skeptical that you could actually get value out of it because basically that social data is so new and it's changing so fast. Do you find the same attributes of the data when it's behind the firewall or is it more stable? Is it more, there's more metadata? Can you get better predictions out of that data? Is there a real schism there? Yeah, well, first of all, that text tends to be more, in the case of English, more regular English that someone's used to writing or dealing with words. You look at things like Twitter and Facebook. The language. Easier to grok, you mean? Yeah, grok, exactly. The language evolves continually in terms of abbreviations and things like that. So it does make it a lot harder to get into in mine in some cases, but the internal information tends to be more contextual, more consistent. There's far less sarcasm, right? Sarcasm is always a big conundrum in the social side because even human beings don't get sarcasm sometimes, right? So it's very hard to tell when something's sarcastic or not. But you don't have that as much behind the firewall. You know, if you have a customer case, you're going to be able to find what products are involved, what are the issues? Are they happy with the interaction? Are they getting frustrated? That's the kind of stuff you can pick out very easily. There's more sort of standards if I can use that term and you have classification engine as well. I got to believe it's easier to classify behind the firewall than it is outside the firewall. Is that true? Absolutely, and in certain industries like financial services, we actually have companies that want to use ontology, so we have an ontology module. If you're in life sciences, there's tons of ontologies out there that lets you create a very nice kind of conceptual mapping of all the structured information and bring it into your decision-making process. How do you guys price? Can you share that with us? Did you price it by module? Yeah, our pricing is actually something that really kind of differentiates us in the market. Our philosophy has been that we don't practice, we call punitive pricing. The idea that if you're successful with our technology, we don't want to be charging you more. As you add more servers, we actually have an application-based model. So someone wanted to use our technology for customer experience management. They would get an application license, and in that license, typically either perpetual software or term-based, they would have unlimited data, unlimited servers, unlimited sources. And the way we scale is through, we believe we're going to deliver business value and you're going to want another application, another application. That's how it's worked out, and a lot of customers are on their third, fourth, fifth application at this point. So when it's application, it's the application that you guys provide or that they develop or a combination of both? We kind of either jointly develop or partners develop with them, sometimes customers develop them themselves. But yeah, so we define what that scope is, generally, of the business problem areas, kind of what defines the scope of the value that the technology's delivering. How do you measure that? Is that, can you automate the measurement of that, or is it just sort of? Another application pops up, here's a bill. It's something that we work with customers, and once you get in a situation, it usually is pretty obvious as to what that scope is and that definition. But something that we work with individual customers, figure that out, because we view it as a partnership. And that's part of the discussion is we want them to be successful with technology, know what the costs are going to be, and be able to clearly measure their ROI and see a big gain when it comes to using our technology. Have there been any applications, Rick, that have really surprised you, personally, that popped up? Well, we've done a number in oil and gas, actually, which is kind of an area where, I didn't really know much about it, but we started getting into it with some of our partners and there's some health and safety is a great example. So when you're operating oil rigs, anytime there's an injury, it shuts the rig down for a certain period of time while they investigate, and they write up reports about what happened, what machinery was involved, and that's all tech space. And right now it goes unanalyzed. It just gets filed away in a drawer somewhere, but if you're actually able to analyze that in real time and send out, you know, proactively take action and say, hey guys, we're having a problem with this part of the rig. There's this part that's faulty, it's causing injuries. If you keep one rig up for one day or it would have been down otherwise, it's millions of dollars. You know, that's kind of cases where you get this huge gain just by getting the right information to the right person at the right time. Right. How about stuff you're working on personally that you're excited about? Yeah, so I'm in the Partners Organization at TVO, so I've been doing a lot of work with some OEM, so we actually can't really talk about too many now, but in the next few months we'll have some pretty exciting partnerships coming out to market. Also working with a lot of the big partners to create some value-added solutions, companies like Censure or TCS, Tech Mahindra, all big partners of ours that I'm working with, and really what we're doing is looking at what are the repeatable solutions that they can bring to market around this unified, rich view of information. So you saw, did you see Nick Silver's keynote? Unfortunately I did not get a chance to go to it. I think he was going to answer the question, when is the big data going to deliver big results, but I had to leave, and I'm not sure he answered it, but when do you think big data is going to deliver big results? I think it's already doing it. We do it every day, so. That's what I thought too. So I was thinking, well, wait a minute, aren't we doing this already? And Jeff Kelly, of course, is quantifying all this, so when do you think big data is going to deliver big results? Oh, I think 2012. Next month, middle of the month, around three o'clock on a Tuesday. All right, thanks, Rick, for coming on theCUBE. It was always good to see you. Keep it right there, everybody. We'll be right back. Our next, excuse me, our next guest is from Altrix. George Matthews coming on. He's the president and COO. They're doing some really interesting things. We're going to talk to him and unpack some of the things that they're doing, which tableau. This is theCUBE. We'll be right back after this word.