 Live from New York City, it's The Cube. Here is your host, Jeff Frick. Hi, Jeff Frick here. We're on the ground at The Cube's fifth birthday party at Big Data NYC. We're in Manhattan for Big Data Week New York City. We just came off our first ever capital markets panel, which was really great. We had full day of Cube coverage today, a full day of Cube coverage yesterday, but we want to give you a little feel for what's going on here at The Cube's fifth birthday party. So I'm joined by my next guest, Shabam Maharish. Yes. And you're the head of financial services and insurance for fractal analytics. Great, well welcome. Thank you, thank you so much. So what do you think of the, you're there for our little panel? What did you think of the panel? Yeah, no, I think it was great. It was, you know, we represent more of the practitioner side of the business. I thought it was great to listen to, you know, more of the technical and the software side of the house and what's happening there. So we are more users of these platforms as opposed to systems integrators. And we represent Fortune 500 companies that use those platforms. So which is great, right? That was just basic messages that the winners are going to be the practitioner. So you're actually out in the field. You have kind of a unique model that we talked about off camera. You're really helping companies implement these strategies and technologies. Absolutely. We have over 700 data scientists that help companies implement strategies. So think of us as a business consulting firm that's implementing machine learning and advanced algorithms to help solve business problems. So that's really the problem. So you solve the problem of there's just not enough data scientists out there? Absolutely, and the whole model is it's an on-site offshore model. So we leverage our teams in India, China, globally to actually solve these problems. So it's a 20% of our team is typically in the US, 80% in India. So that's how we solve the people problem. Right, and is it a consultative basis where you go in and help them do that? Or do you actually implement and run it on an ongoing basis for them? Or do you do that and potentially do a leave behind when you're finished? So it's a great question. It's both. So it depends on what the client wants. But mostly we try to implement an ongoing relationship because that's when we can retain the knowledge, what we learned from the previous gig that we did, a project that we did for them. So that's really the model. OK, and another thing I want to follow up with you is there was an interesting comment on the panel, right, about the flyover states. And the joke that in California, New York, Boston, we're always ahead of the curve and what's happening in the flyover state. But Amy from Cloudera said, you know, I'm down in the weeds. I'm working with companies. So you're out with the customers. Talk a little bit about how this has really gone beyond the coasts and how it's kind of moved beyond kind of POCs and the real production instances. No, absolutely. And I would side with Amy. I mean, I think there's a lot of work happening. I mean, if you think about agriculture itself. I mean, today with Internet of Things, there are sensors on those, you know, factors and, and, you know, the equipment, agricultural equipment, and that's only going to increase. So I totally agree with Amy that there's a lot of work going on there. From our perspective, our whole mission is to help companies institutionalize analytics. So, you know, understand the leverage of data, understand that it's a data-driven world tomorrow. And so that's really the model. And so it doesn't matter where you are, because we can bring that sophistication to you. But then how, but then how do they start to adopt that culture if they don't have, either they didn't have it or they just don't have the people? I mean, is there a level of kind of becoming an analytics-driven company, even though you're not a bunch of data scientists, you know, kind of at the levers of the switches of the business? No, that's a great question, because a lot of our clients are in that spot, right? They, they have a choice, right? If you're in the Midwest, where will you get those PhDs that don't want to work for Google? Right. They still want to work for you as a data scientist. And that's a struggle that they have. But there is a, there's a journey that they have to follow, right? So meaning you could still evangelize analytics by not going all the way to the extreme end of sophistication and implementing machine learning. You could do simple things like visualization, right? If I can visualize your data for you, if I can show you how as an insurer, you know, your claims are trending across the U.S. You can drag and drop, click in, zoom in into a map. Very intuitive visualizations. I can get you to appreciate data and I can slowly evangelize that within your firm and then you'll understand why advanced analytics is required to get to the next level, right? So there's a, there's a maturity curve there. So I know people a lot of times are sensitive talking about customers, but I'm just, I'm curious if there's any examples that you can, that you can share with us either generically or a real customer of kind of an eye-popping experience that one of the customers had when either it was something completely different than they, than they thought or, you know, it really drove home the power of thinking in an analytical way. No, absolutely. I mean, we have a lot of text mining algorithms that we have built over the years. And so an insurer wanted to figure out at the time that there's a claim, right? There's a first notice of loss. So your car gets hit, you get a re-rendered kind of collision, you know, you call in the insurer, a claims adjuster gets tagged to your, you know, case and they file what's called a first notice of loss. So there's a lot of text in there. So can you mine this text that they're filing in their systems to come up with, hey, will this go to litigation? Will I get a third party claim on this? And we have built algorithms that can mine that text to kind of come up with those insights. So that was a really interesting, you know, output of what we did. And what did they discover? I mean, were they able to get ahead of the curve? Were they able to channel it in a different workflow based on those risks? So we were able to classify over 80% of claims and classify them correctly that we were able to predict that these would go for litigation. Using unstructured data that had always been in the system, but somebody had to read it. Nobody actually kind of applied any machine learning to it. And we did, yes, absolutely. All right, awesome. All right, well, Shabam, thanks for stopping by. I know you want to get back to the party. We're, again, we're on the ground here at Big Data NYC for the fifth birthday party or the cube coming off our first capital market event. I'm Jeff Rick, you're watching theCUBE. Thank you.