 from Las Vegas, it's theCUBE. Covering IBM Think 2018, brought to you by IBM. We're back at IBM Think 2018 from Mandalay Bay in Las Vegas. My name is Dave Vellante. I'm here with Peter Burris, my co-host, and you're watching theCUBE, the leader in live tech coverage. Scott Hebner is here as the Vice President of Marketing for IBM Analytics. Scott, welcome back. Good to see you again. Thank you, glad to be back again. So you heard Ginny this morning, you know, very inspiring speech. I love her talk, she's really good in front of an audience and one-on-one. What were your takeaways, specifically as it relates to your group? Well, I think the theme of this whole conference is a lot of these technologies over the years that have been purchased separately and are thought of as separate, you know, quote-unquote segments, are really all starting to fuse together. They're becoming different facets of the same challenge that, you know, a large majority of our clients have. And that is really this evolution towards a more AI-based set of business models. Right, and there's a stack of things that need to be done to make that successful. Now you got to move to the cloud, for the agility of it, the economics of it. You know, you got to get more value out of your data and make your data ready for AI. And then you can start to more effectively train your AI models and, you know, and allow them to continue to learn and everything. So it all really comes together and I thought that's what she was framing of what IBM is trying to do uniquely. Yeah, and I think it came across that way and obviously this conference is about bringing together all the, you know, the separate. And your organization is evolving. I mean, when I think about, when you think about IBM, go back Peter to the, even the Gersoner days, and he said, no, we're not going to split up into a million companies. We're going to have one face to the customer. And then, you know, obviously IBM was very successful there. You now had some major changes in the marketplace and you're responding to those. Yeah, I think that's exactly right. We're being very customer driven. One of the great advantages of IBM is that we have so many customers, right? A mix of new ones, a mix of ones we've had for a long time. And we have so many people that engage. If you think about the size of IBM and how many are engaged with customers every single day at all levels, from the very most technical to the people that manage relationships, we learn a lot collectively. And with all the new technologies, particularly around digital, net promoter score, all these things, we learn a lot about what they're trying to do. And that's what's driving us to fuse these strategies together into a more holistic one. And that's what you heard this morning from Jenny. So I also really enjoyed what I heard this morning from Jenny. It reminded me, though, of one of those television shows where people bring in their old family artifacts and then people price them. And I imagine enterprises today literally looking at their data, the 80% that nobody has visibility to and finding, you know, grandpa's letter from Abraham Lincoln. And using and discovering that this is a source of value that they've never envisioned before. Is that kind of the mentality, the conversation that we're having today? No, that's exactly right. I mean, a large, large majority of CEOs have declared their data to be a strategic asset, but only about 10% of them believe their company treats it that way. And it leads to the statistic that you just referenced, which is 80% of data is either unanalyzed, untrusted, or inaccessible. So they're sitting on a goldmine of data, right? It's not just empirical customer records, but it's increasingly IoT and sensor data. It's behavioral data. There's a goldmine there. So step one is how do you take advantage of that and get more value out of it, right? Just in today's world, right? And then it really becomes fundamental to being successful with artificial intelligence. You have to have an information architecture. You know, we kind of say if there's no IA, there's no AI. But you have to have that information architecture to be successful. And that's really where we're focused on at this conference today, is getting that data ready for AI. So getting the data ready for AI is, there's a lot that goes into that. But when you consider the notion of data as an asset and what we heard from Junior this morning, it seems as though in many respects, there's kind of two models happening in the industry. Let me see if I got this right. Companies that make money off of your data and companies that aren't going to make money off of your data, would you agree? I mean, is that kind of how the split is starting to happen in the industry right now? Yeah, I think that's right. I mean, I think a large majority of our clients are using their data within their firewall to operate their businesses better, better understand their customers. No, I mean something different. Yeah, sorry, I apologize. Companies that are going to make money off their customers' data. Yes. And companies that are not going to make money off their customers' data. Yeah, right? Yeah. Yeah, what I'm saying is, no, I get the question, is different companies have different business models of what they're going to do with their data. Some see it as an asset to run their business more effectively. Others see it as a direct asset that they sell in resell and resell, right? What I'm saying is, the majority of the customers we deal with are looking at their data as an asset to run their business better. And that's the basis for the argument that the incumbency that we're entering back into the era of the incumbency because of all these rich assets are not currently being utilized. Is that right? That's right, right. It all starts with the fact that the data is fragmented everywhere. Business partner networks across different databases. So step one is to make that data simple and accessible. But once you do that, that's not the end of it because you need to make sure that the data that people are using is trusted. You have to have that trusted analytics foundation. So you got to integrate it, replicate it, catalog it, cleanse it, manage its life cycle. You need to have one version of the truth, right? That everyone works off of, which is a major problem, by the way. It's a whole notion of governance and that falls into other categories like privacy and all the compliance challenges that customers have. And then from there, you have that foundation where you can start to drive more insights out of it through things like machine learning and pattern recognition, right? And as you start to build those skills around data science, it starts to get you really ready for that next step on that ladder to AI. And that's where a lot of these customers are sort of figuring out, how do I get on this roadmap to AI? And 85% or so say they're going to get there in the next five years. There was a great study from MIT Sloan that came out last year of 3,000 customers. And it was very clear the difference between the pioneers and that are having success and those that aren't is the pioneers have figured out how to make their data ready for AI. It all really starts there. And that's really what we're focused on here at the show. And let's talk about that incumbent theme that was part of Jenny's talk this morning. And you're right, the incumbents, their data exists in silos. Even though they're maybe data companies like a bank, they're organized perhaps around their products or a manufacturer might be organized around the bottling plant, as you say. Whereas those companies that are AI driven have data at their core. So it's a challenge for the incumbents. How are you helping them get from, close that gap, that AI gap, if you will. Right, and that's exactly what I was just saying before, is that the data is incredibly dynamic and grown at exponential rates. And not only through what you just mentioned, but there's acquisitions, there's different business partners that evolve through your networks, your client data, things of that nature. Your data sources, yeah. Data sources are changing. Then you get into the technical layer of all different types of data, from images to empirical data. And then you get into different databases and it becomes a very heterogeneous mess. So step one is make it simple and accessible. And doing that through big data. And being able to view through a single layer all the data as it changes, right? And if you don't have access to your data, then what are you going to be training your AI algorithms on, right? And again, from there, then you've got to govern it in a way that it's trusted data. This is a huge challenge for customers because they get different versions of data that tell them different things, which is the single version of the truth. It's kind of like if you've ever been on a treadmill, your watch says this many steps, your phone says another number of steps, and the treadmill says the third number of steps. You're like, how many steps did I really take? They have that challenge every day. And so when you get that foundation and information architecture together, then you're ready for AI. And what this MIT Sloan study showed was, bad data is paralyzing to AI. And no matter how sophisticated your algorithmic AI capabilities are, bad data is simply paralyzing. And so that's where it really needs to start. And to circle back to your point about 80% of data untrusted on analyzed and accessible, that's got to be step one on that ladder to AI. So how are we going to use ML machine learning AI to help us get our data ready for machine learning AI? Well, that's exactly what we're doing in the IBM portfolio of data and analytics products is we have this theme called machine learning everywhere. So it actually is in almost every part of our platform. So hybrid data management uses machine language, or machine learning to help do a much quicker assessment of how you bring data together and analyze it and things of that nature. We use it in the governance. In fact, we have a technology prototype that we've been working with some customers on that will do the work for GDPR, the European compliance guidelines, in probably a few days to a week versus months and months and months. So we'll go in and do all the entity associations for all your data to help you organize in a way that you can actually manage what to do with the compliance, right? And then obviously machine language is fundamental to just business analytics in general, pattern recognition, the traditional analytics tools will help you understand the data as it's presented based on what you are trying to get out of it. But often you don't know what you're trying to get out of it. Machine learning gives that data science method of actually uncovering patterns, which you can't really see. Creating models. Yeah, creating models. And do you add the neural networks to it and deep learning? You know, it's really literally a ladder that you're building that when you get to AI, you're going to be a lot more successful as you've built that trusted foundation underneath it. And I think Jenny was touching on that to something to read this morning. And that's what we're majoring on is that that data is really the key element of AI. Scott, who are the roles that you see developing this information architecture, getting ready for AI? You know, CDO, CIO, Chief Digital Officer, where do they all fit? Yeah, I think it leads under the CDO. And actually I would see both CDOs, Chief Digital Officer and the Chief Data Officer in their collection of data engineers, data stewards, things of that nature. Because again, you've got to start by getting that information architecture in place, right? It also involves sort of a new generation of data developers that are building cloud-based data intensive applications, particularly off events-based data, right? Which is a little bit different than customer data from sensors and all that, where you need that massive ingest speeds and it's those data-driven applications in the cloud that are really starting to incorporate machine learning. So they become really key. And then from there, if you think of it as a collaborative life cycle, you get into the data scientists that are applying analytics, they're applying a more sophisticated version of mathematical programming and data science. And then there's a new sort of subset of them which are the AI developers. So it's really from the data engineer right through business analysts, there's a life cycle of people that are part of that team. And they all have to work off a common platform, a common set of trusted data to be successful because it can no longer segment it. And your strategy is to build tooling that allows all of those roles to collaborate, at least maybe not the chief digital and the chief data officer, but the data engineer, the data quality engineer, the application developer, the data scientist, right? And is that correct? That is absolutely correct and the CDOs. So actually what we're announcing at the show is a new offering called IBM Cloud Private for Data. So if you're familiar with IBM Cloud Private, it's our private behind the firewall cloud platform. We're coming out with a new offering that plugs into it, it's based on Kubernetes. So it runs on IBM Cloud Private for Data and will run on other Kubernetes based platforms. And it is a fully integrated data and analytics platform where no assembly is required. And it will provision in minutes a pre-assembled, customized experience for you based on what your role is. So if you're the CDO, you're the data scientist and I'm the data engineer, we're all going to have a different set of requirements of what we want to get out of the data and what we're looking to do. It will pre-provision that for you very, very quickly. And so, and you're all working off a common platform. It's collaborative in nature with dynamic dashboards. So you can see what's going on. It's really taking the building blocks that you need to be move up that ladder and integrating to microservices into a cloud platform that is just lightening fast in terms of, not only its ingest speeds of data, but more importantly, the ability to provision new users. So it's a major step forward in making it so much easier, so much more simple to get more out of your data and to get your data ready for AI. So last question, you have this giant portfolio. We just finished our big data report. You guys, IBM came out number one. Not a lot of that with services, but still you got a lot of software in there as well. You've been working hard to pull those pieces together so that clients can simplify data. Okay, here's where we are at 2018. Where do you want to take this thing? Well, I think again, I think step one is these unified experiences. Because again, we were kind of majoring in this conversation about the desegmentation of how people work in a business, what technology, what data they use. Because with AI, it really does need to come together, right? And so we're trying to do the same thing for the users, which is provision based, almost on demand, what you need based on what you're looking to do. And I think what's going to change as we go through time is it becomes more and more machine learning based pattern recognition. It's more automated and customized and personalized based on what you're trying to do. And that's going to allow businesses to move at a much more rapid pace. And again, I think the overriding theme when you look over a five year horizon is is your data ready for AI? And that's where we're moving this whole thing. It's about the data, it's about the people and their skills, and its ability to move quickly. And that's where the linkage with cloud comes in. Getting to pervasive AI, but you got to get your data house in order first. You got it. Scott Hebron, thanks very much for coming to theCUBE. Thank you. Great to see you again. Great to meet you. All right, keep right there, everybody. We'll be back with our next guest. You're watching theCUBE at IBM Think 2018.