>> Narrator: Live from Las Vegas, it's theCUBE covering IBM Think 2018. Brought to you by IBM. >> Welcome back to IBM Think 2018. This is theCUBE, the leader in live tech coverage and my name is Dave Vellante, and we've been covering IBM Think, this is our second day. IBM's inaugural conference will be here at three days, wall-to-wall coverage. Al Martin is here, he's the IBM VP of Hybrid Data Management, client success, I'm going to get that in there because it's such an important part of the title. Al, welcome to theCUBE, thanks for coming on. >> Thank you, pleasure. >> We'll start with hybrid data management, what do you mean by hybrid data management, what is that? >> Well I think, it starts with data, and they call it information technology not data technology for a reason, meaning I have the pleasure or the burden, one of the two in terms of being able to set up what we call the AI ladder. Meaning you start with data, you push it up the stack, push value up the stack that being analytics, ML, AI, and data today is a challenge, I mean it's a huge problem. It doesn't matter what size client you are, it's a challenge for you, and so it's unstructured, it's structured, it can be in the cloud, it can be on-prem. So when we say hybrid, it's across- The challenge that I have is across all those different foreign factors. We've got to make data simple and accessible all across all those foreign factors, that's hybrid. >> It's a tall job, tall order. >> Pretty much all jobs. >> Okay, how do you do it? >> How do I do it. Well, very carefully. We develop technologies that do just that. What we do at via is common analytics engine first and foremost. We use an engine like, no matter what foreign factors say when I'm in an appliance, I can query the appliance, and then if I want to take that work load outside that appliance and put it against my own hardware, I can take that database out and still query, do the same analytics, I can put that in the cloud, do the same query and analytics, no different. So, the way we do it is we don't care whether it's structured or unstructured, we don't care whether it's no SQL or SQL, we'll do both, we'll do analytic processing, we'll do operational processing and we try to do it within the same footprint, that's essentially how we do it. >> Okay, so what I like about this is your chan is every customer, I mean of every company (mumbles). >> That's the challenge. >> What's the conversation like when you walk into a client or a prospect, what are the words they're using to describe their problems, helps us understand that. >> That is a great question, because it is very difficult to get those words out very often. A lot of clients are struggling where they are on what I call the maturity curve. So, to that point, what I typically do is start with a conceptual maturity curve, and if you can imagine a graph going from left to right, it's a hockey stick a value relative to maturity, and so we figure out where our client is on that maturity curve. By example, imagine four quadrants. On the left-more quadrant is operations, that's your ERP systems, your billing systems. If they're there the opportunity is cost-optimization, or the deal is operational systems don't typically do well with analytics. So if they're looking at analytics then they'll move to the next quadrant and do data warehousing, then the opportunities tend to be data legs, you might want to get into Hadoop, and then once you graduate from there you go into self-service analytics, that'd be like the third quadrant, and then you're thinking about Spark as a common analytics engine, you're thinking about IOT, and then you start getting into machine-learning, and by the time you hit the fourth quadrant, that is where new models begin and you're really driving machine-learning and driving the progress to AI. When I look at that model, those four quadrants I just walked you through, is I'm pushing as much as I can to both the developer and the business, and give them the empowerment, and when you do that then governance comes into play, data science comes into play, new personas come into play. So it's quite a challenge, but I find where the client is on that graph and figure out where they want to be, current state, desired state, and then we draw up a plan to get them there. >> So let's talk about those, sort of. That is I guess the maturity model, right? We started with a core systems, ERP, transaction systems, you started to build data warehouses, data marts, they were largely bespoke systems, it was sort of an asynchronous data move, you have it build big complicated cubes. Still do, still doing that. >> Still do. Still doing it in many cases. >> And they're driving decision support, but it got really expensive, and a lot of times it was like a snake swallowing a basketball to make a change. Okay, so then along comes Hadoop thrown into a data leg like you say, it's got a reduction of investment, but then you got to get value out of it. Now you're talking about self-service analytics, Spark comes into play, simplifies things a little bit and now you get ML, more automation. My question is, as you proceed, as customers proceed down that journey, is there a hybrid data management architecture that has to be put in place so that these aren't separate bespoke pieces that I leave behind but they all come together in an enterprise data model. >> Here's the way I would explain that, in making the complex as simple as possible. We figure out where they are, and then there's essentially five different key elements that we key on. One is hybrid data management, that's what I'm responsible for, and by example, the database we use supports HDAP, which means it'll do both analytical or warehousing and transactional processing at the same time by example. When you're looking at unified governance that would be number two. Unified governance is, the best way to describe that is, is unified governance is done for data, what libraries do for books, same concept. And then the third one is then when you're pushing that closer to the developer, then that's when you get into data science and the models start building upon themselves and that's where the magic happens. Those are the three, but there's two more. Under data science, I usually call out machine learning, because machine learning is very important. I mean that enables that path to AI that everybody talks about, the bridge to AI. And then finally I think a key to any client strategy is open source. Most people don't know that IBM is one of the largest contributors to open source, like a patchy Spark by example. We believe in open source because it increases the pace to market, so if you have those five different strategies, that's how you be successful. Within my organization you can have an appliance, for hybridated management, you can have an HDAP database, we have one-click data movement, all those things go into that to make up that complete solution. >> HDAP by the way is hybrid transaction and analytic processing. >> That's exactly right. >> You see those worlds come together, I remember the Z 13 announcement a couple of years ago, you guys made a big deal out of that, and so that's actually happening is that right? >> That is absolutely happening, yes. >> So that involves what actually doing the analytics in the transaction system, is that right, in the database of the transaction? >> I mean it depends on work loads, there's a lot of depending factors, but yeah, that's the- >> As opposed to what, putting it in some kind of Infiniband pipe into my data warehouse. >> Well you talked about it earlier, where previously you have to create complete separate data marks, you have to transition and use ETL to go from an operational store or a transactional store, to an analytical store completely separate. Trying to do both those in the same databases is our objective, that's HDAP. >> Excellent. Now you're also running the global elite program. >> I am. >> What is that all about? >> Well, let me back up for a second and tell you how we got here. I am running the global elite program but it started out just as a sheer campaign of driving personalization for our clients, pretty simple right? We have got the technology now to really personalize our experience with our clients. Using ML and some of the same technologies that I talked about. By example, we use ML and Watson to both internally and externally with clients, in other words, internally we make recommendations to our analyst, externally you can use a bot and ask them the questions. We're pushing all our content out, essentially free-of-charge, opening it up, we have very aggressive push to push that content out, and we're driving direct to expect. So that's just standard now for us, that's the basic, but then we've taken that further because we want to treat each client relative to their needs and profile, so what we've done is, for the platform offerings that we have, we just came up with a new offering called Enhanced Support. So what that does is it's front-of-the-line service. Consider it your airline priority service, so it's front-of-the-line, it's faster response time targets, and it also provides some consulting, and then on top of that, we've got what's called a premium tier, and that premium tier does everything of what I've already described, but then it adds a named context, and experts, to work directly with you with one foot within IBM, and one foot within whatever client in that expertise required. So I give you all that, global lead is at the top of that. These are our partners that are innovating with us, that are rewarding us with their business but they're innovating with us, they're serving as references, and together we're partnering and transforming together whether it's retail, insurance, or otherwise. So those are a small set of our global elite clients, and I encourage any clients that are listening out there, if they feel like, hey I want to partner directly with IBM, I want to push the envelope, references are in my future, I'm in. >> What are some examples that you can share with us? >> What we've done, we tend to have a motto with the global elites that we never say no, and I'm still waiting, I haven't said no yet, but we'll see if that ever comes. Well we never say no, and what we've done by example as an evolution of the global elite program is think conferences like this, a lot of times you can only send so many people. So what we've done is we've taken a mini conference, and we call it Analytics University, and we've taken that directly to clients, and we'll do a day or two and do this conference in a miniature scale focused on the areas and the content that they prefer. The other thing we've done is then a lot of times when we do that, we'll find interests and visions that they have that they have not been able to really get into a road map or progress. So then we'll bring them into the lab and we'll do design thinking sessions, and then we'll work together. And in terms of doing the design thinking sessions, what we essentially, ultimately accomplish is one independent road map between two different companies, because they help set our road map, we help influence theirs, and all of a sudden they've got a strategy to the future, and it's organically aligned with ours. >> Excellent. Alright Al, let's put the bumper sticker on IBM Think 2018, it's only day two here but what's your takeaway from the conference. Trucks are pulling away, what's the bumper sticker say. >> The bumper sticker says, make data simple. >> There you go. >> That's where my head's at, make data simple. I got a podcast out there that's called Make Data Simple. I'd encourage everybody to listen to it, we get into all these different technologies, but I think we make data simple with a- The wider the breadth we get data we can drive value up the stack. >> So, Make Data Simple podcast, right? >> It's actually under Analytics Insights in iTunes. >> Analytics insights under iTunes. >> That's all me. >> Alright, beautiful. Yeah, Make Data Simple podcast, Google that and you'll find it. Al, thanks very much for coming to CUBE. >> Alright, thank you. >> Pleasure having you. Alright, keep it right there everybody, we'll be back, right after this short break.