 Welcome back to theCUBE, live here at Hadoop Summit in San Jose. We are towards the tail end here, day two. It's been a great couple of days. We've had some really great guests on, and that continues now. We're not done by a long shot. So we've got Jay Duff, CIO of L2C, joining us here in theCUBE, along with Will Daly, Client Solutions, from Chiseta, our good friend, I'll be founder of Chiseta. We had on a little earlier today, talking about analytics and the financial services business. So why don't we start, Jay, talk a little bit about your company, what you guys do, and then let's just dive into kind of applying analytics, big data analytics to the financial services. Okay, L2C is a well-established company. We calculate credit risk scores for individuals, consumers. Where we're unique is that we specialize with the underbank in particular. So obviously you're familiar with FICO scores. There's almost 100 million people in this country where FICO scores don't really provide the full picture of their credit risk. So you can imagine students, immigrants, people with some deep blemishes in their credit history. L2C looks at alternate data sources from a variety of areas, a variety of dimensions, and we use predictive modeling to provide something that begins to look like a FICO score. But again, it is applicable where a FICO score doesn't provide sufficient data. So what allows you to do that? Is it big data? Is it bringing in more data sources? How do you kind of go beyond what the FICO score is capable of? What's changed is with Hadoop, now we're able to look at unstructured data and bring unstructured data into the algorithm. So in the past, we would purchase data. It was always highly structured industry standard types of formats. Today, we can look at unstructured data, and more importantly, we can match it with the structured data. That's something that we weren't able to do in the past. The other thing that's interesting is, today we can trend it. So in the past, we would typically look at attributes for an individual at a specific point in time, and today with Hadoop, we can look at trends, we can look at accelerations. For example, we can look at home values, and it changes in home values. So in the past, you might look at someone's credit profile, and make a determination based on the value of their home, or the value in a neighborhood, whereas today we're looking at the trend for that neighborhood, and making much better use of historical data than we were able to in the past. Interesting. So in terms of the opportunity here for financial services, we'll maybe get your take as well. I mean, what are the big, this is a great use case, but from a big picture, what are some of the fundamental possibilities here when you're applying big data to the financial world? I mean, we know what the economy's been going through recently in the bank meltdown a few years ago. What, how can we, how is big data helping address that, and how are you guys helping to do that? Well, with the clients that I've been working with, I would tell you that the most immediate interest actually is in trying to understand their storage and their storage solution. So can big data replace tape? That's a very common set of questions that we answer. At the next level is actually trying to map out and understand their actual data. A number of companies recognize that they're very dependent upon their data, but they really have no overarching data strategies. And understanding and mapping out that data and be able to put it into a single source. And I think Jay, this is one of the things we accomplished. Really, we looked at all the different sources you had and actually finally gave you a single view of all that. And that's incredibly powerful. So just even those simple things like that, let alone actually getting into the determining how to add business value. And I think that's another thing that we're doing for L2Cs. We really see some significant future business value. In fact, you made a comment the other day to me, potentially this could do what for your business long-term? Well, long-term we see that L2C is going to be able to provide much more differentiated solutions. Today are able to provide different models for different credit profiles or credit applications. The credit risk associated with someone that's going to open up an account for a cell phone is much different than a credit card, which is much different than a used car. So our models are flexible as a function of the underlying loan type. What are some of the examples of these new unstructured data sources that might surprise people that actually have an impact on figuring out risk for a potential consumer? Some of the new things that we're beginning to look at revolve around education levels and employment levels. There's affiliations, you'll find information in press releases. There are complexities in being able to use unstructured data and unstructured data, and that's primarily where my focus is. There's also a lot of complications with privacy and there's federal regulations that dictate what you can and can't use to determine credit scores. So certainly we're very, very sensitive and making sure we're compliant with that. At this point, my challenge as a CIO is more can I take unstructured data and match it up and make it structured such that I can run it through a model. But again, these other non-technical issues are just as important. Right, absolutely. We don't hear a lot about security and privacy some shows like this, because we're focused on the analytics, some of the really cool things we're doing, but I think definitely, especially when you're dealing with sensitive data like you are and not just separate data points, but then as you combine them, you're creating a whole new set of really sensitive data. So yeah, how important is that to you, the security, the data you're complying with regulations? I mean, it sounds like it's pretty much key to your business. Well, it's a federally regulated industry, so clearly there has to be legal review and our customers, large lending institutions, are also very sensitive. They have their own compliance and they will audit us in detail to make sure that they understand the sources of the data and that we comply with all of the aspects of the law. One of the things that's interesting is we're beginning to get into aspects that the law really didn't consider. We can see the intent of the law, we can understand what was, the law was a mental accomplishment, and so we have guidelines, but it becomes complex as you start to look at the data sources that weren't considered in the past. Great, let me add a comment to that. One of the ways that we really need to think about this is in terms of asset value. Data has tremendous value, but if you can't protect it, you can't secure it, that value goes to zero. So there's more than just the regulatory, there really is a business driver for why you need to manage the security, the access, the authorization to get to that data. Very powerful, very important. Many of the clients I work with, they consider that data to be one of the most valuable assets they have on the board, so this is a significant issue the community has to address. So kind of going above and beyond potentially what the regulations are, because the regulations are sometimes behind the time, so to speak. So are you finding that your customers are looking to go even above and beyond what's required of them? Absolutely. And it's in terms of how they use the data, again, to provide value to their business, just to give you a practical example, social security numbers. There's a lot of types of analysis and reports you can run that you don't need access to those social security numbers. On the other hand, people from the HR department would absolutely need those. Our value is that we can take all that data into a single source, right, and then provide different views into that data versus the old way of building the data for HR, another one for marketing, right? But if we can't solve that problem of guaranteeing that only the HR people see those social security numbers, we've got a problem. And this is a very key driver that we need to solve in the community. If we're going to hit the business community the way we intend to. Very interesting conversation. Security, analytics, yeah. I mean, security, like I said, it really, it comes down to, you know, topic that doesn't always get a lot of coverage, but I think it really needs to in this context. When data, as you said, is such an asset. So thanks so much, great conversation around analytics, security. There's, you know, a lot more we can talk about. Thanks so much for being on theCUBE. We appreciate it. We'll be right back from Hadoop Summit live here in San Jose. Thank you.