>> Announcer: Live from Las Vegas, (electronic music) it's theCUBE. Covering IBM Think 2018. Brought to you by IBM. >> We're back at IBM Think 2018. This is theCUBE, the leader in live tech coverage. My name is Dave Vellante, and this is day two of our wall-to-wall coverage of IBM Think. We've been doing IBM shows for years. This is the big, consolidated show, 30 to 40 thousand people, too many people to count. Cameron Clayton is here. He is a GM of Watson Content and IoT Platform at IBM. Thanks for coming on. >> Thanks very much for having me. >> So quite a show, right? Standing room only! >> A large, large show. >> Standing room only and also great announcements. >> So tell us about your announcements. >> Yeah, so we got to couple of things we're really, really excited about. The team's been working really hard on for the last few months. One is a way to train Watson to make Watson even smarter than it already is out of the box. And so, we've been building data kits by vertical industry. So for financial services, for travel and transportation, for the hospitality industry, for health care and for government, on how do you give Watson a high machine IQ right out of the gate as opposed to having to train it in your area of industry. And so, once again, we're really focused on making Watson the AI system for Enterprise, and this is another step on that journey to make Watson really, really smart. >> It's really prioritizing it in a way that's much easier to consume. >> Much easier to consume, and if you think about it, there's a lot of jargon in each industry, right? To be an expert in industry, you got to know a lot of jargon, understand the context of that. An AI system doesn't know that unless it's taught that. And so we are teaching Watson that. And then how to apply it successfully in each of those industries. So it's a pretty material leap forward in how we're training Watson. >> So it hits the content component >> Cameron: Hits the content. >> And then industries you're knocking down? Where are you starting? >> Yeah, so we're starting with financial services. We're launching in travel and transportation and in hospitality. So we're basically, this is a pretty fun one, I love food. But basically Watson went out and scanned the entire internet and collected all the recipes that it could find on the internet and trained itself on food. And so, you can ask it now questions about food, what restaurants, about really specific things. If you're a vegan you can find out what's available near you. If you're gluten intolerant, you can find out things on the menu like that. But then there's other things, like in the travel and transportation industry. Virtual agents for travel agents, they can ask questions of Watson, and it can ask very specific, very deep things, very much like a human would. And so you can say a simple thing like, "Where should I stay in New York?" And a human would respond, "Well, are you a member of any hotel rewards program?" Normal AI chatbot wouldn't. It would just say, "These are the lists of the 4,000 hotels in New York." Watson will actually ask human-like questions to give you the best answer possible. But all that requires training, and that's what were built in with these Watson content data kits, and we're really excited about 'em. >> So I'll come back to that. But so if I take that example of Watson Chef, there's this discussion on AI for the enterprise versus AI for consumers. >> Right. Are you crossing over? That was kind of a consumer-y application. >> Cameron: Yeah. >> Is that just an example? >> It's just an example. No, it's very much about AI for the enterprise, right? And so the four priority industries that we're focused on, first is financial services, sort of the sweet spot for IBM. The second is supporting our government clients to make sure that Watson is trained in the language and nuisances the of government. The third is Watson health, so the health care industry, both the regulation and the language itself. So everything from pharmacology, et cetera. And then the fourth is travel and transportation. So it's very much about making Watson the smartest AI system for enterprise. That's absolutely its focus. >> What's the IoT angle in your title? >> Yeah, so-- >> What's going on there? >> I run the IoT platform for IBM, and so The Weather Company, which is how I joined IBM, which I also run, really is one of the largest IoT platforms in the world, which was actually a big part of the acquisition case for acquiring The Weather Company. We're now bringing the ability to ingest 35 to 40 billion data requests every day with The Weather Company platform to the IoT platform. We've combined those things together. So we can ingest data and content at a scale unlike pretty much anyone else in the world, sort of second only to Google in terms of the scale of data and content we can ingest. And we use that data to help train Watson on one hand, and on the other hand, to support our clients in multiple industries around the world. >> Yeah, I remember when IBM did that acquisition, Bob Picciano told me, "Well, you got to understand. "This is an IoT play as much as it is a data science play." So how has that evolved, come together, with IBM's core? >> Yeah, so I think in a couple of ways. One is, it's taken the way the company was mostly a domestic US business. IBM, in the last couple of years, has globalized that business in a very material way. A great example is in aviation, where we have the top 30 US operators. Now we have hundreds of operators all around the world helping them make decisions every day. At its core, this IoT platform that started with the way the company is now much larger than that, has grown into a decision platform, right? We make recommendations for people to make decisions. Mostly that's with Watson and AI, but sometimes it's just with machine learning and more traditional methods. >> So you got some other stuff going on. >> We were talking off camera >> We do. >> about this real-time closed captioning. I was showing you our video clipper tool. You said, "Hey-- >> Yeah! >> "We have something very similar." We're going to maybe talk and see if we can't-- >> Yeah, that'll be great. >> collaborate. I can't wait to try that out. So talk more about what you're doing with real-time closed captioning. It's a mandate, >> That's right. >> for broadcasters and other folks like YouTube. >> That's right. . How are you helping them? >> Yeah, so, as you mention, closed captioning is a regulated space for broadcasters, both local and national. It's a cost center for them, right? They have to do it, and it takes time, people, effort, and energy. We're automating that and we're doing it in a real-time way, so in true real time. So as we're speaking, Watson is listening. It's recording and it's annotating everything that goes on in the video clip. And then it's also breaking it up into essentially a highlight reel, right? And so you can ask questions. Hey, show me the highlights of the US Open or the Masters Golf Tournament. And it'll automatically select the very best clips that came from that tournament based on sentiment analysis, tone of voice, trending key words that were showing in social media, and surface those clips up, typically to a human editor who will then process them. It basically automates a system that today requires human intervention to deliver and makes it completely seamless by being in real-time. >> So Watson will analyze social data, Twitter data, take the fire hose and say, "OK, based on the Olympics," or whatever it was, "this is what was hot." >> Cameron: That's right. >> Curling was off the charts hot. >> (laughs) Curling is always hot in Olympics. >> Hashtag curling. >> Right. >> OK, cool. >> That's right. >> And this is a product that's out on the market today? >> It's a product that's launching here at Think and is being tested by multiple clients right now and is a really great accuracy, quality scores, 95% plus accuracy. But most importantly, it's no human intervention. So no person has to do anything, and it meets all of the regulatory requirements. For digital content creators, which are the fastest growing part of the video ecosystem, people like yourself and others, are also using it to automatically meta tag all their clips. So not only does it do sentiment analysis of the clips and the content itself using the closed captioning, but it's also going out and measuring social media key words and hashtags that are trending and looking for those key words in the closed captioning and clipping that out and surfacing it to make it easier. >> And I consume that as a monthly service kind of thing? >> Exactly, exactly, yep. >> How 'about GDPR? That's hot topic these days. Can you help me with my GDPR problem? 'Cause the clocks ticking on my defines, kicking in. >> Clocks ticking on GDPR. If you haven't started on GDPR yet, you're in some trouble. >> You're way late. >> You're way late, but you better call IBM pretty quickly, and we'll parachute in and try and help. >> How can you help? >> So I think we can help in multiple ways. So one is, obviously, our services group with GBS. We're doing thousands of engagements trying to help people with GDPR. I think, secondly, is we've got a big effort with our consumer weather business to be ready for GDPR. We have 250 million users of our weather app around the world, and they'll have to be compliant here pretty quickly. And so, we've got that all set up, ready to go. And then, these data kits also learn the regulations, right? So you can ask questions of Watson about GDPR and your specific use cases as a customer, and we'll show you how to apply the regulations of GDPR to your business. >> So earlier on, you talked about these data kits. I mean, in my head I was thinking SDK. >> Cameron: Right. So how does that all work? >> Yeah, so you can, you basically on a SAS basis, you essentially rent these data kits, everything from a general knowledge kit to a industry specific kit for financial services, to a sub-industry like wealth management within financial services. And you basically can rent each of those pieces. Within the government category, we have a GDPR capability, along with other regulatory capabilities within the data kits. >> OK, so how does that work? I sort of train my internal system? >> It's super easy. You, basically, go to Bluemix, and you can just use it as a subscription out of Bluemix is the fastest, easiest way to do it. Secondly, you can talk to any of your IBM associates about how you use data kits with Watson. It's always used in conjunction with Watson services themselves, is how you basically deploy our products. >> Let's say I got data all over the place in my organization, it's siloed out, and I'm freaking out because I've got personal data on an individual here and one over her and one over here. What do I do? I point my corpus of data at Watson, and it helps me extract from itities, dedupe, surface? >> The first step in all of our engagements is to listen and understand exactly where all the data is, and everyone's on a journey, right? From on prem to hybrid to some public cloud and everything in between. >> Dave: And they don't know where it all is. >> And they don't know where it all is. And so, step one is for us to go in and listen. We have a rule in our group, two ears and one mouth, use them proportionally. And so we go in and we try to listen, find out, map out sort of a architecture of where our client's data is. And then understand what problem they're really trying to solve because, often times, there's lots of good ideas, but there's only a couple of problems that really matter to that client to solve. Right now, GDPR is certainly one of those problems. But whether it's revenue or efficiency, we can help, but we really need to understand what the problem set is first. And so we have an engineering team that goes in and does sort of architectural work and listens upfront. And then we go into a sort of solutioning mode to solve problems. >> One of the question's we often ask on theCUBE is, how far can we take machine intelligence? How far should we take machine intelligence? What are the things that machines can do that humans can't? How is that changing? How will they complement each other? How will they compete? You must think about that a lot in your role. You're augmenting, sometimes replacing a lot of human tasks. But what are your thoughts on those big picture questions? >> Yes, I think we've, as a company, work really, really hard to make sure that we are always augmenting people wherever possible. We fundamentally believe that every job is going to be changed by AI, but we believe that humans are really good at creativity, at curiosity, and at risk management. We don't really think about us being good at risk management, but from when we're born, just learning to walk is a risk management exercise, right? Look at any toddler wobbling, learning to walk, you sort of realize it's a risk management exercise. AI systems have to learn all these things. And so surfacing and recommending decisions is what we believe Watson and AI is best equipped to do, and then have a person actually make the final call. >> Great. All right, Cameron, hey, thanks very much for coming on theCUBE. >> You're welcome. >> It was really a pleasure meeting you. >> Absolutely, likewise. >> And look forward to the follow up. >> Absolutely, we'll follow up. >> Excited to see that. All right, keep it right there everybody. We'll be back with our next guest right after this short break. You're watching the show theCUBE live from IBM Think 2018. We'll be right back. (electronic music)

>> Announcer: Live from Midtown Manhattan it's theCUBE. Covering Big Data New York City 2017 brought to you by SiliconANGLE Media and its ecosystem sponsors. >> Hey, welcome back everyone live here in New York City it's theCUBE's special presentation Big Data NYC. This is our fifth year doing our own event here in New York City, our eighth year covering the Hadoop World ecosystem from the beginning. Through eight years, it's had a lot evolutions, Hadoop World, Strata Conference, Strata Hadoop, now it's called Strata Data happening right around the corner. We run our own event here, talk about thought leaders and the expert CEO's, entrepreneurs. Getting the data for you, sharing that with you. I'm John Furrier co-host theCUBE with my co-host here Jim Kobielus who's the Lead Analyst at Wikibon Big Data. And Chuck Yarbough who's the Vice President at Pentaho Solutions part of Hitachi's new Vantara. A new company created just announced last week. Hitachi in a variety of their portfolio technologies into a new company, out to bring in a lot of those integrated solutions. Chuck great to see you again, theCUBE alumni. We chatted multiple times at Pentaho World, going back 2015. >> Always he always great to be at theCUBE. >> What a couple of years it's been. Give us quickly hard news, it's pretty awesome you guys have a variety of things at Pentaho you know with Hitachi, that happened, now the market's evolved, what's this new entity, this new company they're bringing together? >> Yes, so the big news Hitachi Vantara. So what that is, two years ago Hitachi Data Systems acquired Pentaho and so fast forward two years. A new company gets created from Hitachi Data Systems. Pentaho, in a third organization at Hitachi called the Insight Group so Hitachi Insight Group. Those three groups come together to form Hitachi Vantara >> What's the motivation behind that. I mean, I go connect the dots but I want to hear your perspective because it really is about pulling things together. The trend this year the show is as Jim calls it, hybrid data, integrated data. Things seem to be coming together, is that part the purpose? What's the reason behind pulling this together? >> Yeah, I think there's a lot of reasons. One of them is what we're seeing not just in our own business, but in our customers business, and that is digital transformation. Right, this this need to evolve So Hitachi Vantara is all about data and analytics. And a big focus of what we do is what Pentaho's been doing for years which is driving in all kinds of data, big data, all data. I think we're getting on the cusp of closing out the big data term, but you know, it's all data right. >> Data everywhere, every application. >> And applying analytics across the board. One of the big initiatives, part of why Pentaho was originally acquired we were actually Hitachi Data Systems was a customer of Pentaho when we got acquired, so we we knew each other pretty well. And part of the reason for that acquisition was to drive analytics in around internet of things. The IoT space, which is something that Hitachi being a very large IT and operational technology, OT, company probably does as well as anybody if not better. >> So going back couple of years, I'm just looking at my notes here from our our video index. You visited theCUBE in 2015, but really the concepts have evolved significantly. I want to just highlight a few of them. What data warehouse optimizations, we talk about that. Data refinery concepts, 360 view as applied to big data. Again that was foundational concepts that all are in play right now. >> Absolutely. >> What is the update in those areas? Because refinery, everyone talks about data refinery, you know, oil, the easy oil example but I mean, come on, data is everywhere it is most important, you can use it multiple times unlike oil, as you were pointing out. >> So interesting you bring that up. So to me data refinery in a digital transformation really in an IoT world where lots of data is is streaming through in fact, yesterday I read something by IDC that 95% of all data in the future and the data growth is dramatic it's 10x what it is today in just a few years. 95% of the that growth of data's IoT related. The question is how are you using most of that, right, and what what are you going to do with it. So that data's is streaming through, there's a lot happening, we can do things at the edge, we can apply analytics and filtering and do things. But ultimately that data is going to land somewhere and that's where that refinery, think of it as the big data center refinery, right, where I'm going to take that large amount of data and do the things that Jim does, you know and apply machine learning and deep algorithms too really. >> I had some thoughts on the IoT Jim and I were arguing, not arguing, discussing, with others in theCube about the role. >> We were bickering. >> The role of the edge because I was saying the refiner of the data can come back depending on what kind of data or you push compute to the edge, kind of known concepts, people been discussing that. But the issue is been, how do you view the edge? I'd love to get your reaction to that question because a lot of people are saying you have to think of IoT as a completely different category, than just cloud, than just data center, because the way some people are looking at IoT I know this can be semantics whether it's industrial or just straight internet of things device, or person, that is a different animal when it comes to like what you call it and how it gets put into a bucket. I mean most people put a lot of the IT bucket but. Some are saying IT edge should be completely different category of how you look at those problems. Your thoughts on how that IoT conversation shape. >> The question I always ask when I'm talking to somebody about the edge is, well what do you mean? Because it is something that can be defined a little bit differently but in an industrial IoT context I think, you know we look at it as one, you you have to know what those things are you have to really understand them. And part of understanding those things is having a digital representation of what those things are. >> A digital twin? >> A digital twin. Right, or asset avatar, as we call it at Hitachi. >> Oh I like that. >> So this idea of really managing those assets, understanding what they are and then being able to know what the current state, what the previous state, things are like that are. And then that refinery we just talked about is sort of where that information goes to so you can do other kinds of analytics right. But when you're talking about the edge, typically what we're seeing is the kinds of analytics might happen at the edge, are probably more around filtering you know, it's not quite as complex of analytics that's what we're seeing today. Now, the future I don't know. >> Sort of tiered analytics from the edge on in with more minimal, I mean, not minimal that's the wrong term, with a more narrowly scoped inference. Like predictions and so forth being handled at the edge with larger more complex models being like deep learning whatever being processed in the cloud is that it? >> Yeah that's exactly the way that I see it. Now the other thing about the edge, depends on who you're talking to, again, but what is an edge device or the the gateways or the compute right, so part of IoT is in my mind, it's not cloud, it's not on-prem or it's not, I mean it's a little bit of everything right, it depends on the use case and what you're operating. We have a customer who does trains as a service in England, in Europe, and so they don't sell the trains anymore they actually manufacture trains, and they sell the service of getting a passenger from here to there. But for them, edge is everything that happens on those trains. And tracking, as a digital representation, the train and then being able to drill down deeper and deeper, and you, know one of the things that I understand is one of the major delays for train service is doors opening and closing or being delayed, so maybe that comes down to a small part and the vibration of it and tracking that. So you've got to be able to track that appropriately. Now, on a train you might have a lot of extra space so you could put compute devices that have a lot of power. >> What's interesting you said the edge, in this context, is everything that happens on that train. In other words, it sounds like all the real world outcomes that are enabled, perhaps optimized, by embedding of the analytics in those physical devices or in that entire vehicle that is essentially. One way that you're describing the edge which is not a single device but as a complete assembly of devices that play together. Amongst themselves and in with the services in the cloud. Is that a logical sort of framework? >> That's why I said I usually ask what do we mean by edge. If you've got millions, thousands, whatever, devices out there feeding sensors whatever feeding this data, collecting, processing you know there's some some level of edge computing gateways, processes that are going to happen. >> Well, my question for ya, I'd like to get your thoughts, as we, again we're having a, we love the hyperbio we think its completely legit and it's going to be continued to be hyped because it's obvious what you see with IoT standing on the edge. But lot of customers we talked to are like, look I got a lot going on I got application development I got to break out my security got to build that up. I've got data governance issues, and now you throw in IoT over the top. They're like, I'm choking in projects. So they they come down to one of a selection criteria. How do they define a working IoT project? And the trend that we're seeing is that it has to do with their industrial equipment or something related to their business. Call it industrial IoT, because if they have something in their business, say trains, as a critical part of what they do, that's easy to say let's justify this. Everything else then tends to go on the back burner, if they don't have clear visibility of what their instrumenting. That's kind of weird do you agree with that? Do you see a pattern as well as what customers are doing by saying I'm going to bring this project in and were going to connect our IoT. >> That's exactly what I see. Industrial internet of things is where I see the biggest value today when you have trains or mining equipment or you know whatever. >> John: Whatever your business runs. >> Your manufacturing line right. and being able to a fine tune those lines to either predicts failures, maybe improve quality. Those are those are impactful and they can be done right now today and that's what we're seeing is kind of the big emerging thing. IoT's interesting to talk about, the reality is it's really digital transformation that we're seeing. Companies transforming into new business models, doing things significantly different to grow into the future. And IoT is an enabler of that. So you're not going to see IoT everywhere today. >> The low hanging fruit is where it gets to the real business. >> Yeah, but it's going to go across all verticals, right, no doubt. >> So what solutions does Pentaho have for digital twins, or managing digital twins, the objects, the data itself, within and IoT context, is this something you're engaged in already? >> So within the Hitachi Vantara, the larger company. Bigger company, we have, we have what we call our Lumada IoT Platform and in that there is this asset avatar technology that that does exactly what you're describing. Now I'm going to throw quick plug out if you don't mind. Pentaho World in a couple, in about a month. >> John: theCUBE will be there. >> theCUBE will be there, and we're excited to have theCUBE and we're going to we're going to give you complete information about asset avatar with all the right people. >> There's a movie in there somewhere I could feel it, Avatar two. There's a lot of great representations of data I want to get your thoughts on how the new firm's going to solve customer problems. Because now as the customer see this new entity from you guys, Vantara's been doing real well, we covered the acquisition and you were kind of left alone Pentaho was integrating in, but it wasn't like a radical shift. Now there's some movement, what does it mean to the customer, what's the story to the customer. >> You know I think it's great news for the customer because Pentaho's always been very customer focused. But when you look at Hitachi Vantara the wealth of technology and expertise. Everything from all of the the great IT oriented stuff that Hitachi Data Systems has done and been well known for in the past still exists. But this broader focus of taking data and processing it in a variety of ways to solve real business problems. All the way to orchestrating machine learning in applying algorithms and then with the Hitachi. >> What specifically in Hitachi is coming into this? Because again this is again a focused solution company now with data, so Hitachi Data Centers, >> Yeah, so Hitachi Data Systems, think of it as the the infrastructure company. Hitachi Insight was the really focused largely on the IoT platform development, with some Pentaho assets and then the Pentaho business. But here's the thing about Hitachi, very large company, builds everything. Mining equipment and and all kinds of stuff. So nobody understands how all those things fit together better, I believe, than Hitachi. But some of the things that we have at that organization is this idea of the Hitachi labs. And data scientists that are really doing interesting things Jim you'd love to get more embedded into what some of those things are, and making that available to customers is a huge opportunity for customers to now be able to embrace a lot of the technologies we've been talking about. I said last year that this year was going to be the year of machine learning. And if you look through the expo hall that's what everybody's talking about. Right, it's AI or machine learning. >> I'm wondering if you're commercializing R&D that's coming straight out of Hitachi labs already or whether the Vantara combination will enable that. In other words, more innovation straight out of the labs, into into the commercial arena. >> That's something that we are absolutely trying to to, right because there's great things that these lab organizations and at Hitachi they're big labs. They're really legit, I kind of joke about that. The kinds of stuff that they're able to bring about now, Pentaho is part of the engine to help actually commercialize those things. >> Chuck I know you're looking forward to Pentaho World I'll give you the final word here in this segment how you see the big data worlds evolve. Take your Pentaho hat off and put your industry guru hat on. What's happening, I mean this AI watch, that's pretty obvious, not a lot of blockchain discussion which is going to completely open up some things we getting on the decentralized application market which is going to compliment the distributed nature of how we see a date analytics flow and certainly the immutability of it's interesting. But that's kind of down the road. But here you're starting to see the swim lanes in the industry, you've seen people who've been successful and the ones who have fallen by the wayside. But now the customers, they want real solutions. They don't want more hype, they don't want another eighth year of hype, they want OK let's get into the real meat and potatoes of data impact to my organization, call it digital transformation. What's happening, what is going on the landscape. >> So you know I mentioned before and to me it's digital transformation which is a big huge thing. But that's what companies are interested in that's what they're beginning to think. If they're not thinking about those things they're falling behind, five or six, seven years ago we talked about the same exact thing with big data. It's like a big data is really you know it's a big opportunity and they're like well I don't know those that didn't adopt it aren't necessarily in a position now to transform digitally and to do some of the things that they're going to need to evolve into new business opportunities. >> And the big data examples of winner is the ones who actually made it valuable. Whether it's insight that converted to a new customer or change an outcome in a positive way, they go that wouldn't have been possible without data. The proof points kind of hit the table. >> That's right the other thing is you know, who's going to win, who's going to lose. I think people that are implementing technology for technology's sake are going to lose. People that are focused on the outcomes are going to win. That's what it is, technology enables all that but you've really got to be focused on. I want to get your quick, one more quick thing, before we go I know we got we're tight on time but I want to get thoughts on the open ecosystem. Open source going to whole other level. The projections are code will be shipping at an exponential rate, it's be a lot of onboarding of new stuff, so open obviously works, community models work, partnering is critical. So we're seeing that good partnerships, not fake deals or optical deals or Barney deals, whatever you want to call it. But real partnerships. You starting to see technology partnerships. What's your view on that, how is the new Vantara going to go forward, are you going to continue to do partnerships and what's the strategy? >> Yeah I think the opportunity with one, Hitachi Vantara is we have a breadth that can touch many different aspects. So as Pentaho we had great partnerships, very meaningful but it always comes down to what we doing for the customer. How are we changing things for customer. So I'm not a believer in those Barney kind of relationships those are nice but let's talk about what we're doing for customers. >> Yeah, real proof points. >> You guys will continue to parner. >> Yes, we will continue to do that. >> Okay great, Chuck, thank you so much. CUBE coverage Live in New York City in Manhattan it's theCUBE with Big Data NYC, out fifth year doing our own event in conjunction with Strata Data. Now bless the new name of the show. It was Strata Hadoop, Hadoop World before that. But we're still theCUBE covering eight years of the action here back with more after this short break.