 Live from Orlando, Florida, it's theCUBE. Covering Pentaho World 2017. Brought to you by Hitachi Ventara. Welcome back to theCUBE's live coverage of Pentaho World. Brought to you of course by Hitachi Ventara. I'm your host, Rebecca Knight, along with my co-host, Dave Vellante. We're joined by Rene Wegelen. He is the IT manager of VikeHouse. Thanks so much for coming on theCUBE, Rene. Thank you for having me here. So VikeHouse is the Finnish national bed agency. The wholly owned by the government. Tell us more. Yeah, we used to have like three companies. Now we are merged as a one and we operate every money gaming thing, all the money gaming in Finland. So that includes like the from casino to lottery, to strats, tickets, sports betting, horse betting, whatever that is. And we gathered money, of course, pay out some good winnings as well. So, but everything we make under the line that goes to the good causes. And I mean everything. And you are IT manager. So what is, what are your responsibilities? Yeah, responsibilities like the developing whole of the, well, all the IT things we have from architecture to doing the IT procurement development, enhancing the how we work. So the public policy on betting is, hey, let's have a single state run monopoly and we'll take the winnings and put it to the public good. Makes sense. And is there any competition from internet, for example? Yeah, in the internet we are like, it's like a full competition. Although we are like a legally based company in Finland and we operate and sell only to Finnish people, the people itself, they have all the freedom to choose whoever they want to play with. So in that sense, it's full competition and have been so for many years. So you have to have great websites, great customer experience, user experience, competitive rates, all that stuff. And okay, so, and good analytics. And good analytics, yes. That industry is obviously very data heavy, always has been. So how do you use analytics and data to compete? Yeah, so we have been doing the product analytics for quite a long time and then we established the customer ship. So in the Finland we have like a 5.4 million habitats and we sell only for the 18 plus year old people. And at the moment we have more than 2 million registered customers already. So you can imagine that we have that vast amount of data from the customer. And we use that data, for example, to promoting the services, promoting the games, targeting, making some recommendation. We build our own recommendation engine, for example, and utilize all that kind of data. But as you know, the gaming is also, it's like a two-edged sword. There's a happy side, but there's also the dark side. So it does cause the problem. So we try also to use the data so that we want to identify the bad patterns when somebody is about the loose control of gaming. So we use also the same data that we want to seize, for example, for these players, we want to seize all the activities of marketing, for example. We don't want anybody to get into problems because of gaming. So that's a really interesting tension here, is that you obviously want to make money in this, but you also have to watch out for the finished society. And as you said, if there is a compulsive gambler or an addicted gambler, you need to act. I mean, is that... Yeah, yeah, that's really big part of our responsibility. And if we didn't have any data or if we couldn't process it fast, we couldn't know who is problematic gambler and who's not. Since vast majority is, of course, it's enjoying. It's a nice habit. Play a game of poker every now and then or go to the casino for once or twice a month, for example. But then there's the small portion of people who we want to protect that so that they don't get into the debt. That's not our intention. And the level of protection that you provide is you stop marketing to them, is that right? It's not like you intervene in some other way. Yeah, of course, we want to promote that. If you want, you can stop and close your accounts or these kind of activities. So you promote cutting the cord, basically, saying, okay, let's... So instead of marketing, we say that this might be a problem to you, so... Yeah, let's take a break. Yeah, you should take a break. So as Dave was saying, I mean, you're really, because you are competing with private entities, you really have to have a great interface, great customer experience, great rates. How much does this put Vikehouse really on the vanguard of this kind of technology more so than what other government agencies are doing in the sense that you really have to stay really on the cutting edge of these things. Yeah, we have to be like a back. So how much do you then, you talk to the health agency or other government agencies about what you're doing and sharing the best practices about capturing customer attention? We are actually talking more to the new players out to the field who already live and breathe through the data. So that's where we can learn. And I would say that we are also the leader, also in the lottery area itself, but also in quite many other industries as well. So we have been doing this for a while. So we have had the luxury that we have already gathered some experience and opened some parts and then, well, maybe learned also from the hard way how to not to do it. So we of course didn't succeed it in the first runs, but you just have to go and have a trial and error in some areas as well. So. And you have multiple data sources, obviously. You maybe talk about how you're handling those data sources, are you ingesting, how you ingest those into Pentaho, what you do with it, how you're operationalizing the analytics. Where does Pentaho fit in that whole process? Yeah, Pentaho, we use that as like the ETL process. So to get these 360 view of the customer, we have like a various data sources. After the merger, we like a triple the amount of different sources. And I think more than quadruple the amount of data. So of course, just to make the data and the work of the analysts easier, we need to make some transformations to the data. And in that area, the Pentaho has its place. And in the future, and what we are also expecting like the future versions to help us with is to like the tag in the more real-time data. So for example, we can put in the real-time data feed for the one physical place so they can see like which machines are used well, which are not, or is there any other activities that they can learn right in their place? So are you in the process of instrumenting the machines at this point? Yeah. And so you're putting, how does that work? Is it a rip and replace? Is it a some kind of chip that you put into the machine? How do you instrument the machine? It's a good thing so that, I mean, we have like actually we design our own slot machines. Okay. Even so we can like build up from the ground up. Design it in. Yeah, we design it that the hardware supports like they are big IOT machines but also the software will support us. And then you've got connectivity. Is it hardwired? Is it physical or is it a wireless connectivity? We use, well, whatever is available. So. Depends. Yeah, yeah. And when we are developing like new type of games, for example, when the slot machines should have like online all the time like jackpots available, then of course we have to think about like what's the quality of service of the network as well. But so far we have been like using whatever is available. So what does the data architecture look like? I don't know if you could paint a picture. So you've got the machine, let's just use slot machines as an example. So you have the slot machines, you've instrumented those, you're doing real time analytics there and maybe talk about what kinds of things you do there and then where does the data go? How much data do you persist the data? Maybe talk about that a little. Yeah, so we get like the slot machines and other sources as well and have like a Kafka Hadoop area where we collect everything. Then there's Pentaho doing the ETL work and we store all the data that goes through it to the Vertica. So we have HPE Vertica there and in that Vertica there are like lots of users, they're like SAS analytics use that and the Hadoop as well. So then we have some reporting, financials, finance department, they also utilize it. But then we are also like building up some new things like Apaches Kudu is one thing that we want to set up there just to make the life of analysts much more easier. So they are at the moment having a little bit hard time in some areas how to utilize the data and especially like how to use the different analyst tools from different cloud vendors for this data since we are still at the moment in on-premise. So everything is on-premise partly because of the government requirements. So some part of data they require that we keep it in within the Finland. Right, so can we call that your private cloud? Can we use that term? It's not private cloud yet. But it's your on-prem data set. Yeah, we are going to that side. Someday it will be your private cloud. Okay, so you have the edge device, which is the slot machine. And then do you send all the data back into Vertica or no, probably not, right? I mean, does you? Not yet. But it will be, yeah, it will be. Of course we have to make some decisions like what data will be important and what is not. So not all the data is valuable. But especially when it's like a connected somehow to the customer or the retailer as well, that data we also keep more than a year. So we are not doing all the analytics just for the short time of data, but also want to seek out the long trends and make new hypotheses out of it. And the Vertica system is essentially your data warehouse, is that right? Or okay, and then are you doing sort of, well you mentioned recommendation engines, so you're doing some form of, it's a form of AI as far as I'm concerned. Are you doing that? Where are you doing that? Is you doing that in your data center and is that another layer of the data pipeline or is that done in the? Yeah, it's done partly on-site, but also on the AVS. So we use Amazon services in some areas where we can use those. So the recommendation, for example, and part of the Vekas.fi, that's part, some blocks are also on the AWS. So it's a three tier, right? So there's the edge and there's the aggregation at Vertica and then there's the cloud modeling and training that goes on. And Pentaho plays across that data pipeline, is that right? Yeah, it's one major player in our data platform in this sense so that it will take care of quite of many different kind of transactions so that we have like the right data in the right place. Cool, all right, I'm done geeking out. All right, so Rene, before the cameras were rolling, we were talking a little bit about the difficulties of cultural change within these organizations and you were talking about something that you're working on in Finland that's not necessarily related to Vekas. Can you tell our viewers a little bit about what you're doing? Yeah, we are also setting up like a teal Finland. So just promoting this like a next phase of organizational, well you cannot call it belief but vision and perspective. So we want also to promote these kind of activities. So I know that especially with the big data movement you have also seen the cultural changes so that not the normal organizational ways of working are not just are not efficient enough so you have to liberate the data. You have to give like the freedom how to use the data, what kind of hypothesis, what kind of activities are done and these cultural changes also with the teal movement it's like getting like a next big leap. So this is, well it's a side project but it's also really heavily work related. And how open is the Finnish tech community to these ideas? I mean, is there an adversarial relationship within the people who don't necessarily welcome the change? I mean, how would you describe it? I believe it's really open. We have already I believe a handful of companies who work and who operate by these, from this perspective and more is popping out and we are establishing one cooperative like to support this movement and maybe to create the new spin-offs which can be for profit. All right, let's get to the heart of the matter here. How do I beat the house? I knew you were going there Dave, I knew it. Just between us. Obviously I'm kidding, but different games have different odds, right? I mean, and those are, you're transparent about that. People know what they are. What are the best odds? Is it slots, the best chance of winning? Or poker? Yeah, slots is good side and also whenever you go to Cassie, it has a top notch, so 90 point something. Right, right, of probabilities. Yeah, but of course I have to say that the house wins eventually. So you guys, the bookies always win. Right, exactly. Okay, but so the higher the probability the lower the payout and reverse, presumably, right? So the lottery? Lottery, Euro jackpot if you, yeah. Low odds, but telephone numbers if you win. Yeah, yeah. But Dave, you can't win if you don't play, okay? I'm just saying, I'm just saying. Every week there's somebody who wins. Right, right. So I cannot be me or you. Or me, or me, I mean, maybe, okay. So what do you do to the guys that count cards? You like break arms or you put them in jail? It's been lived, this is now, come on. Nobody does that. Do you have, do you? Of course, yeah, that's probably something we could, in the future, also, to use data more efficiently than we use at the moment. So that's one part, like how people behave versus machines behave. So for example, in the online poker, the card counting program, that's one problematic area for the industry. Right. Are you working with behavioral finance experts in this to sort of understand people's behavior when it comes to this? Yeah, we work, for example, with the psychologists to understand this and the same goes with problematic gambling as well. So you have to know about how people behave. And do you have customers outside of Finland or is it pretty much exclusive? No, it's, sorry, it's exclusive club. You have to move to Finland. I can't get in. You know, you have to move to Finland and then we welcome you, yeah. Awesome. He's going to immigrate, I think, any day now. Yeah, but hey, it's one of the best countries, so. Thank you so much for coming on theCUBE. It was a lot of fun talking to you. Thank you. Thanks. I'm Rebecca Knight for Dave Vellante. I'll have more from Pentaho World just after this.