 Live from Las Vegas, Nevada, it's theCUBE covering EMC World 2015, brought to you by EMC, Brocade, and VCE. World 2015, day two of three days of wall-to-wall coverage. We've been coming to EMC World for, I think, six years. This is where we started theCUBE in 2010, so we're always happy to come here. We're changing it up a little bit this year. We like to talk about innovation. We want to be a little innovative. So we're actually going to have a little customer panel. So on my left, immediately, as you all recognize as the dean himself, Dean of Big Data, Bill Schmarzo, CTO of Global Services for the Big Data Practice. To his left, I'll make sure I get this right, is Elizabeth Fletcher. Welcome, CPA and CA, Deputy Director, Smart Leadering and Infrastructure Program at BC Hydro. Welcome. Thank you. It's a big, important job. I like it. And then next to her, we have Chris Pershing, founder and CTO of Eagle View Technologies. Welcome, Chris. Thank you. Thank you for having me. Awesome. So, first off, why don't we give a little background about what you do in your role and the company? Certainly. I work for BC Hydro, which is a Crown Corporation located in British Columbia. We generate, purchase, distribute, and sell electricity to people in British Columbia. My role in the smart metering program is I lead the program currently to develop and implement smart metering technology as well as the other infrastructure, including the application for detecting a theft of energy, which is what my focus has been. Excellent. And Chris? I'm Chris Pershing, Eagle View Technologies. We've got two main lines of business. The first is the capture of very high resolution, very widespread capture of aerial imagery. Both downward looking and oblique, kind of angular looking. And our second line of business is basically extracting content out of those images. Originally, Eagle View's main line of business was taking these images and creating three-dimensional, very accurate, authoritative models that we could use for estimating a bunch of parameters. And we then expanded out, extracting different kinds of content out of those images. All right. Great. Well, welcome. You guys here at Big Data, so what's a Big Data panel? So let's jump in a little bit and talk about what does Big Data mean to you? Are you guys using Big Data? How have you started your journey? Obviously, you're moving down that path. Elizabeth, to start with you. How things change with the growth of Big Data, the explosion, this opportunity all around data? We started our pass land down the Big Data trail back in 2012. We had almost completed our installation at 1.9 million smart meters. And we were starting to receive all of that data every day, hourly information about the energy being consumed on the grid. We needed to do something with that. And in particular, our first use case was for an application that would help us to identify where electricity was being stolen off of the grid. Very significant problem for us. We identified back in 2010. We were seeing at that time $100 million of energy theft every year. $100 million of energy theft. I was going to ask you, my next question is going to be, is that a big deal? But you already answered my question. Yes, so as part of the smart rearing business case, we identified the theft detection benefit was estimated at $700 million Canadian. You estimated at $700 million? That's based on a present value over 20 years. Okay, okay. And was that really one of the main drivers of doing the smart meters? Or was that just one of many benefits? It was one of many, but it is the most significant benefit by far. Wow. How were that grabbing power? They just tapping off and running it somewhere? It's a variety, but it goes from tampering at the meter, tampering upstream, connecting into the grid upstream of the meter, connecting it at the transformer, or even in the primary at high voltage. Lots of different ways that theft can be stolen. $100 million. Yes. It's a lot of Tesla. It's a lot of Tesla. It is a lot of Tesla. Yeah. And Chris, you talked about kind of the evolution of your business for this really high-res imagery and what you've been able to do with it, and it sounds like even modify your business. Talk a little bit about the role the big data played and how you got there from here, or got here from there, I should say. So again, for us, for our company, big data really means the fact that we're capturing and recapturing every couple of years, you know, upwards of 90% of the population in the United States. We have a time history going back, you know, every couple of years over a lot of this area, and we also have the ability to then take these images at ever-increasing resolutions. And the higher the resolution, the more applications you can unlock, more things you can see on these images. And so the analytics we can run on these images, the types of models that we can build from them, the content that we can see and extract and observe from that is really kind of the foundation of what big data means for us. And then to be able to catalog that in a property-centric kind of way so that we build up content and information about particular parcels, we can see how they change over time, we can see what's happening on them. And as that parcel-centric database grows, there's additional layers of analytics we can run on that, both serving the markets that we serve today and, you know, once they're yet to be realized. And who are some of your customers that are using this data? So I've got a couple of bands of customers. One is insurance companies, which, you know, if there's a storm that takes out or damages, you know, could be thousands or tens or even hundreds of thousands of roofs. We have the imagery of what used to be there, what it looked like before, so we can measure without sending somebody up on a broken or damaged roof. So we help them settle claims faster. We help them do it in a safer way. And we help mediate that conversation between the insurance company and the contractor that's got to replace the roof. The other group is the contractors themselves, not just for storm damage, but for, you know, roof swear out. They need to get replaced. They get damaged for other reasons. People have extensions put on, et cetera. So contractors that deal with anything that has to do with the roof and, you know, in more recent years with walls and siding as well. So talk a little bit about, as you're going down this path, right, you know you're going to get more data. You're getting more pictures, higher resolution imagery, but at the same time, so you've got increasing storage requirements and complexity. At the same time, you know that it's going to open up some new business opportunity. Talk about how you guys are going through kind of the business trade-off decisions and making the commitment to go forward. It was it kind of, we know we're going to have something, you know, we're going to have a market for this. Let's just go get it because we can, or was it really opportunity-driven that said if we have this higher level of resolution, then we can sell that particular service or a new service? It's a little bit of both. There's the natural progression of the, you know, the silicon, the sensors getting higher densities. The various optics you can put on the plane and whatnot. So there's a natural progression that can bring down the resolution incrementally, which is kind of the natural flow of it. But then there are those leaves where you say, well gee, if you could get twice the resolution, what else does this unlock? What can we see at this next resolution level that we can't see today and where are the markets that are going to want to know about that data? So both you, Elizabeth and Chris, are much further along down the path of big data than many, many of the customers. If you look at where you are, how you've gotten there, are there any sort of learnings that you could share with other people, sort of the good, the bad and the ugly about things that you've learned that you'd like to share more broadly? I think for me, the understanding that while there may be technical challenges, the biggest problems may tend to be in the area of governance and making sure that the appropriate business groups are involved and are in fact driving to the solution not being led. They need to drive business. Interesting. So you found that if the business leaders were leading the initiative, you had a lot more success and staying power for the initiative underway. Yes. Yes. Are those business users also helping you to uncover sort of your next use cases coming on the path? Absolutely. And we're finding that there are business users coming to the project team to cry out for access to the data. They want the access now, please. And what can we do to make sure that they get it as quickly as possible? That's a great sign for the organization that you get that kind of hunger and adoption of data. Chris, what about your business? What have you learned along the path? I guess I learned that to the extent that you can recognize ahead of time the value of the data that you're generating and kind of look ahead to see where you might be able to use it and repurpose it and get additional value from it. You know, to really recognize that potential, the more lead time you have to try to organize it and try to structure things in such a way that you're in a good position and take advantage of those analytics later on. You know, we're in a very high growth mode from our get-go, and so there's a lot of scramble, scramble, scramble. And to that extent, I guess we've got to rewind a little bit. If you had a little bit, if you're able to step back and think about how you would catalog and organize this and how you'd structure the approach, you know, it just saves a little bit of headache down the road. You seem to be in a point that a lot of organizations want to get to, which is the opportunity to create new monetization opportunities. How are you working with the business in order to sort of identify and realize those opportunities? We've got really good relationships with many of the insurance companies and with many of our large contractors that we work with. And so with that relationship, we can ask them, you know, we can work with them. What are other processes that go through, especially manual processes? What are the other reasons why you come out to the site in the first place? And are there additional questions that we can answer for you before you get out there? It's not always obvious what the value of that data is going to be, right, and how you can potentially monetize it. But there's always this kind of innate feeling that more is better. And now with a lot of the technologies, you can store more. It's not, ah, what do we get to store? What are we going to get read up? So it's interesting as that develops. So let's switch gears a little bit and talk about kind of surprises. As you started on this journey and the data started coming in and you started to actually put some analysis on it, were there any surprises that you just had no idea that were coming? I think we realized fairly quickly that it was going to be more complicated than we thought. And in that the complexity of what we were trying to do was more than we could consume in a very short period of time. So we ended up making it into multi-phases so that we could start with the first phase being relatively simple analytics, get our feet wet, come to understand it, develop the processes, and start to build that data scientist knowledge so that we could then move on to the next phase where we added more complexity. And we're now at the process of implementing our third phase in this fall, which gets us to where we want it to be all along. And is the complexity a function of more data, more variety of data, different ways to analyze the data? How do you see complexity and value unfolding over time in these kind of phases? For us it was some of the advanced algorithms that we were trying to apply, some of which we had patented, as well as incorporating additional data sets to in particular allow us to estimate where we have missing data because you don't always have all the data. Sometimes you have to estimate it, and that's important to be able to do that with accuracy as well. How about you Chris, any kind of fun surprises? Well I think one of them is that, as we grew up so to speak, as our volumes increased and we did more and more business, we were on a big data path before I think we recognized we were on a big data path. And so it was a pleasant surprise to know that we were more right than wrong and kind of the things that we were doing to set ourselves up for the next step. So in that respect, you know, lucky habits to answer. But the other one, with respect to our customers too, is one of our ongoing challenges and surprises is really trying to figure out what they want to do with the data because it's, you know, in a lot of ways it's a new way of looking at things and so it's sometimes they don't know what they want or they don't know what's possible so it's kind of this push and pull of trying to help them figure out what is possible. Now another kind of part of the big data story that's evolving is, you know, you always had your own internal data but now there's all these external sources of data and before, you know, a lot of enterprises did incorporate some of this external data into their own processes to get additional value. Are you guys using any external data sets that maybe you didn't use before that now you can more easily integrate into your process? We incorporated visualization layer using Google Earth because we needed to be able to see where our assets were in a geographic space and we've also incorporated weather data as a means again of helping us to estimate when we don't have the actual data. Has that been provided as a whole another level of insight? It has, I mean the visualization was key to our solution because much of the insight is gathered from the location and the specifics of where a particular anomaly is being identified. I've got a friend who's got some really high resolution photos you want to move off the Google Earth? Yeah. And how about you Chris? Are you using other stuff beyond your primary data that really augment or are you interested in that? Well there are some data sets that help us do what we do, you know, collateral information that the insurance companies for instance want but what we're finding more frequently is we're having conversations with folks that have data sets that they've developed in their line of business that can answer part of the question that they're after and with our imagery and things that we can get on the imagery we can answer part of the question. Another conversation is can we conflate those data sets to gain confidence and have a more confident answer on what they're after and so that's a really interesting opportunity before us and that's one of the areas that, you know, it's one of the paths that we're heading down. Right. What I find interesting about this is in both cases one is that in order to be successful you've got to make sure you're biting off on opportunities on the analytic side because it's small enough so you're not trying to swallow the whole ocean but you've got to sort of realize it's an iterative process, start small, build on success and Chris in your case sometimes your users, customers don't know what they don't know, they don't know what they need and so there's very much a high exploration kind of a process that needs to take place almost an envisioning of the realm of what's possible sort of process. Very great stories. So we're getting the hook. They're going to start the concert here behind us, they've got the electric cello and some other fun instruments so I want to give you the last word. A lot of practitioners like to watch a queue. They love to hear from their peers, right? They're sitting there, they're getting started, they're contemplating getting started, they're trying to figure out how to, they're going to get some investment and sell some people inside it, what would you call them, the hippos that aren't necessarily buying into this whole big data thing. We've always done it this way. So advice for your fellow practitioners if you're looking at them saying, here's what I would tell you based on my experience to help you be successful on your big data journey. Let's start with you, Chris. I guess really think about the potential of what your data might be used for, you know, years down the road and plan an architect for that goal, right? Because I think of, by and large, a lot of data that's being accumulated is being squirreled away and people don't really understand what they could do with it or if there's any value left in it. Right. Elizabeth? I think it would be a let the business drive it. They need to be the one making the choices, making those decisions, picking the priorities of what you need to focus on so that they can make the decisions that in there in the best interest of the business as a whole. Bill, you're out there on the firing lines every day. What they're saying is the truth. It's encouraging this and you can see why they're being successful because I think they're focused on letting the business users drive it. They're trying to iterate quickly and small, iterations of drive success. You're trying to figure out what you could do with the data, visioning what's out there. Both of them are great stories. Congratulations. Thank you. Thanks for coming on, Chris. Elizabeth's sharing your story. Good luck. I hope you got 100 million dollars installed in electricity. That's a lot of electricity. Dean, as always, good to see you. Thanks for stopping by. I'm Jeff Frick. We're at EMC World 2015. Coming to the end of day two, three days of wall-to-wall coverage. We'll be back all day tomorrow, Wednesday. So tune in for more great interviews. You're watching theCUBE. See you tomorrow.