 Live from Boston, Massachusetts, extracting the signal from the noise, it's theCUBE, covering HP Big Data Conference 2015, brought to you by HP Software. Now your hosts, John Furrier and Dave Vellante. Welcome back to Boston, Massachusetts everybody. We're here at the HP Big Data Conference. This is theCUBE's third year at this event. It's an event that really focuses on the practitioner view, a lot of customers here. John Furrier, my co-host, and I are really pleased to have Joseph Kaliel, he's the CEO of Inovo, a company that focuses on the telco industry. Clickstream data, Joseph, welcome to theCUBE. Thank you, glad to be here. So we were chatting off-camera about the history of Inovo. Take us through sort of the comm-score piece, the spinoff, if I can use that term, and how you guys came about and really what you're focused on. Sure, sure. Well, Inovo is a new company. We started it recently, but the business is mature. The business we're running is very mature. We started it back in 2005. The product, the subscriber analytics product that we're selling today. And then we got acquired by Comm-score back in 2010. And we spent five years with Comm-score evolving the product and integrating it with Comm-score technologies. And early this year, Comm-score decided to focus on their core business. And they decided to divest this business unit. And myself and few of my senior managers decided to do a management buyout. And that's when we started Inovo. So we took over the subscriber analytics product line. We have all the customers that are using it. We have today, we have a global customer footprint in North America, Europe, Asia. And that's where we are. We still maintain close relationship with Comm-score. They're a key partner with us. We leverage one of their technologies that's a dictionary used for categorizing websites, mobile apps, devices. It's key component of our platform. And then Comm-score, yeah, keep maintaining that. And we have exclusive licensing to use that in our platform. So talk about the telco industry and their use of data. Very data intensive, data rich. Talk about the industry, how they're using data, how things are changing and where do you fit? Sure. Recently, telco operator, wireless operator started deploying deep packet inspection probes within their networks. Trying to understand really, to get an idea about the traffic that's being generated on the network for many reasons, most of it is technical to be able to manage and upgrade their networks, manage the performance of the network. And also that data contains a lot of valuable information about how customers are using the network. And as we all know today, customers, there are many services available to customers with the advent of smart devices, tablets, and now with IOTs, a lot of things being used on the network. And operators are having the challenges of, well, what kind of services do I need to offer? What kind of rate plans do I need to structure for all of these devices? And how do I provide care? How do I support all these users on the network? And that's where we come in. How do you answer the question or how do people talk about deep packet inspection from the evil perspective? Oh, they're going to probe and throttle my bandwidth. I won't be able to get my Netflix or hey, privacy issues. I mean, deep packet inspection is a double-edged sword. Could you share how you manage that, the balance of that, how people are talking about this? Exactly. Privacy is a big concern and we're very sensitive to that. Our deployments, it's all happening on premises within the network. So none of the data leaves the operator's network. They're in charge, in controls of it. We employ various levels of encryptions to the data, so none of our people can go and query and be able to know anything about any subscriber. So we're very sensitive to that and we've solved that problem using these two main techniques. But the data is valuable. So once we mask the data, then there's a lot of intelligence to be derived for. So we offer a care component and the marketing component and the network component. So the care component is used by the care organizations when you call to report a problem, whether it be a coverage problem or problem with the apps you're using on your device, or whatever it may be. Our software is used to quickly tell you what's really happening at the glance that to tell the care agent that whether you have a handset issue, whether it's a network issue, whether the apps you're using are having issues. All of this we derive it from that. And for us to achieve this speed is using HP, HP technologies like Vertica, because this is really key component of our platform. Talk about the progression that you see in big data adoption in terms of use cases. As we say, the low hanging fruit you pick first. What's the progression of adoption in terms of really knocking down and delivering value immediately? That means that people take a more aggressive approach. Some are a little bit cautious. Take us through the progression. Sure, there is a misconception today in the industry about what is big data and how do you solve or how do you say that you really have a big data solution? A lot of people think that if I deploy the Hadoop cluster I've solved my big data problem. We don't believe that's the case. There are more to it than just having a Hadoop or some similar technology deployed. There is the need to deploy infrastructure to host all these data silos. Today an operating, a wireless operator has thousands of data silos. I heard this that this morning in the general session that there are probably one operator here and they also have over 5,000 data silos. A big data is being able to bring all of these together to host them into an infrastructure like Hadoop, but on top of that you need to build your analytical layers and that's where we come in. We come in, we leverage the data that you aggregated, that you brought in from the different silos you have and our analytical platform brings it together, stitches it together and produce intelligence. In our case, to help care groups, marketing groups and network groups. So yeah, so that was a stone breaker comment. He said the average organization has 5,000 data silos, operational systems, and of course it was interesting. He was saying he apologized for being a SHILTA for HP and then he started shilling his new company, Tamer. That's right. Which is what they do, they take all this variety and they bring together all these diverse data sources. But so how are you doing that? Have you essentially written a component that does a piece of software that does that? Are you doing sort of ETL offloads and reinforcing it? Maybe talk about that a little bit. Sure, for example our care components, we talked about DPI data. That's one source we ingest. We ingest many other sources. We ingest data from trouble ticketing. We ingest data from CRM, from the CRM platforms, from customer complaints platforms. We ingest data about the rate plan. So a lot of different data from the network. We ingest engineering data from the network. Now all of this data should reside in a big data infrastructure like Hadoop. But a platform like us, we bring it all together and stitches it and report it at the subscriber level. So now you're getting a 360 view of what the subscriber is doing, what services they're using, what kind of quality of service they're receiving. And then if they are not receiving good quality of service, is it the network issue, is it the device issue? So we put it all together so that it's easy for an analyst or for an executive to look at it and make a decision. How does that ripple through to the way in which your customers operate? Because we've all made the phone call and you get transferred to somebody else, you get transferred to somebody else, you get disconnected, you got to call back. So that 360 degree view presumably means no matter which services I'm buying, somebody at the other end can see that. Can see it all, yeah. And that's the value that you're bringing. Talk about that a little bit. That's the value we bring with our care module. Yes, when, if you call and they're using our software and on the screen they pull up, they could see all the services that you're using. Within, they could see it within the last hour up to a year ago. We store data for up to 13 months. So they could see it. And then we have a customized report design so that the agent doesn't have to go and drill, try to figure out what's going on. It tells them basically what's going on so they can make a decision quickly. All our reports are designed with, we call them call to action reports. Meaning you look at the report and there is an action you have to take immediately by looking at it. So that's what we do. And to give you a use case in the call on the care side, when you're on the phone with the care agent, you want that call to be as short as possible. And with our software, we've demonstrated that we were able to shave minutes of that call because the agent has everything in front of him to know what to tell the subscriber and how to resolve the issue. And that translates into dollars. And you know how many calls they get every day and if you shave a few minutes from every call, that's huge savings. To the bottom line. So talk about your relationship with HP. What is HP to you? Are they a technology supplier? Are they a trusted advisor? What's your relationship there? They are a technology partner with us. So we use Vertica. We use HP Vertica. It's been working great for us. It allowed us to deliver, to minimize our hardware footprint because of the great compression that they offer. So they offer compression, they offer speed. So we want to be able to ingest, we have to deal with the velocity issue. We have to ingest a lot of data and do it quickly and make that data available for reporting fairly quickly. And then when you report on it, you want your reports to render quickly. So it helps us in all these three different cases. So I mean, as a CEO, you're in the technology business. I mean, technology is fundamental to your revenue stream. So, but talk about how you view sort of data as a component of that technology. How decisions are made as to initiatives or projects that you choose to fund. Sure, sure. On our side, or how decisions are made on your side specifically? Well, we do what we believe our customers are, they want or are asking us for. To give you an example, like on the marketing side, for example, we have the marketing components that we have key customers here in the US and Europe that are in Asia that are using that. And they want to be able to craft a plan, to be able to target their customers for new services. Because they don't know what they don't know what their subscriber are doing, or what services to target, what subscriber group, what subscriber segment. So on our side, we have a data analyst team, data scientist team. And actually the latest term I've heard is data artists. Data what? Data artists. So we've been investing, that led us to invest into creating new models, new modeling techniques, new methodologies to help us really better understand what subscribers are doing. So when the operator wanna do a marketing campaign, they know how to target. So they're not doing the shotgun approach. They know you like streaming, and you know this is the app that you use the most, and that's causing it to exceed your rate plan, then let's target you with the product with the rate plan specific to your use case. And that's what we deliver to them. And for us to do that is a result of our modeling techniques and the models, statistical models that we have embedded into our platform. And you deliver that as a packaged app? Yes. Okay, so I mean this is, I'm glad you sort of clarified that because there's a clear trend. Analytics historically has been very customized. You're an example of somebody who's taken domain expertise within an industry, and packaged analytics and intelligence into an app that I can buy, it's a solution. Is that a, first of all, is that a fair characterization? And is that a trend that you see in the industry? It is a fair statement. What you get today, if out of the box, from when we deliver our platform, is really the results of years of R&Ds and us engaging with customers and telling us what they want, tweak this, tweak that, and we capture all of that. And if you're our next customer, you're the beneficiary of all of that work. You get it right out of the box. Now, every new customer, they want also to do things a little bit differently. And we give them that capability. We have very advanced segmentation engine. We have advanced ad hoc reporting engine that they can use and even build on top of what we give them out of the box. So it's been working great. It's been working great for everybody. So we keep, we try to let everybody gain. Everybody gains from everybody's experience in using the platform. Now, how about the decision probably I asked you before about your internal decision process? Maybe the better question was your customers. You know, Robert Young-Johns was sort of joking, tongue in cheek that you have all these Hadoop projects spinning up all over the place, but what's your strategy? Well, we have this Hadoop cluster. What's your strategy? And a couple of years ago, that was the strategy, is let's try it, see what happens, see if we can get any value out of this. How is that conversation changing in your customer base, the telco industry? How is the decision process changing? Is it becoming executive level, board level, still middle management, still geeky? What's the situation? We're seeing it happening at the management level. The senior execs, we deal with, they understand they need the big data solution. And now they're empowering their managers to really put something in place. And we're seeing this. We are seeing this across our customer base. The decision is going down to the manager, senior manager. But it's an industry that's highly entrenched in its existing processes and for a lot of reasons, good reasons. It feels like it's going to take a fair amount of time for those organizations to become truly what people call data driven. Is that, do you see certain organizations within your customer base breaking out, maybe startups disrupting? What's the sort of landscape there? The Uber of telco, for instance. They all realize they have a lot of data to deal with. And we see a lot of really struggling, what is the right technology for me to use? I think they all realize they need that platform that can host all their data. So it becomes a matter of what is the right technology for me to use. And we're seeing them using some using HP technology, some using different technologies. And then wherever we come in, we try to influence. Now remember, we don't play at the lowest level, at the Hadoop level. We come in as the analytical layer, which they could use whatever technology they want to use for their data. And then our stuff will sit on top of it to do the analytical modeling for that data. I think that's critical because the disruption in every industry is front and center, whether it's Facebook, WhatsApp, Skype. You've got ESPN the other day, it's called under question, investors are concerned, right because everybody's watching Netflix at night and not turning into cable anymore. So these are big forces within the industry that you serve, over the top programming. So I would think analytics is a linchpin of responding to understanding what's going on to be able to offer services. Exactly, exactly. Understanding, you mentioned the over top apps that people are using like Skype now. You use Skype a lot, now you're not using your minutes on the phone. I mean, Snapchat's got the stories, they're putting all ESPN on that, they got mobile devices, the cube, we're over the top, we're not over the top in cable yet or the telcos, soon the contracts are coming. That's right, that's right. That's what we provide, that's what we provide to them. They know what are the top OTT applications customer, subscribers are using. We have the customers, they're using our reports to know who they want to go and cut deals with for these OTT providers. Yeah, and we're actually looking at, we have so much content on the cube because we're on the web, but we have so much programming, we could easily cut 30 minutes to an hour of content for a week. So that's it over the top, so we would have to then, who do we go to? Who would we go? We have to go to a provider, give them the content and then they would distribute it. That's right, that's right. And you're starting to see plans to say you have this much gigabytes per month, but if you use Facebook, it doesn't count against that. Or if you use this, it doesn't count against that. So all of these different crafty plans and tariff plans, all of these, to make that, you don't make that decision just in a word, you need the tool to help you make that decision. Yeah, there's a lot of interconnect issues on the carrier side, right? So like, Google Fiber, that's what's that going to be? You got Facebook putting more video out, they claim them YouTube, now they're autoplaying, but they could be a provider. So as we start crossing in the app, can bolt on distribution. So there's a lot of conflux in this. So how do people making sense using data to be competitive? Or what's your view on all these new dynamics? Again, totally new world. Facebook doing video, Twitter vines are hot. I mean, got Snapchat. That's right, that's right. Now, all what we can do is we tell the operators, here's what's happening. Here's what's happening on your network. Here are where your subscribers are watching these videos. Here are the websites, the types of websites that they are using. So we give them these to arm them with the decision that they need to make. Like we're seeing now also in our space, a lot of traffic being offloaded to Wi-Fi. What's happening? Like what do you do on Wi-Fi that you don't do on the wireless network? Are you using different apps there than you're using? So all of these questions that you need, you need a tool like this to be able to help you come up with the answers on how do I deal with this? And then how do I make sure my subscribers are getting what they need? Okay, so final question. What's the outlook for the next couple years? Okay, telco, big data over the top, all the convergence, the databases are changing, everything can be instrumented. What's going to happen? How do you see the future evolving? That's the million dollar question. You need, I think there are still problems to be solved in big data. There are a lot of problems on the ingestion side of the data that still needs to be solved. We're seeing a lot of streaming. You need data that handles a lot of the streaming data. And then you need the data is going to keep growing. So what's going to happen next year? Like every two or three years, data is going to be doubling. So you need the platforms that can keep up with that data growth so it can help you ingest data fast and it can help the analysts get to that data fast. Now that's where an HP is playing a big role in that in all the solutions that they're offering, like the heaven and the Kafka streaming and all of that. This just keep to be, these technology just have to keep improving to just deal with the influx of the data over the next year. Joseph Khalil helping the telco industry transform through data in Ovo. Thanks very much for coming to theCUBE. Really appreciate it. Sure, thank you. All right, keep it right there. Everybody, John and I will be back with our next guest. This is theCUBE. We're live from Boston, the HP Big Data Conference 2015. We'll be right back.