 Live from San Francisco, it's theCUBE. Covering IBM Think 2019, brought to you by IBM. Welcome back to Moscone, everybody. You're watching theCUBE, the leader in live tech coverage. It's day three of our coverage of IBM Think at the newly renovated Moscone Center. I'm here with John Furrier. I'm Dave Vellante. Interpol Bindari is here. He's the global chief data officer at IBM, longtime CUBE alum. Interpol, great to see you again. It's my pleasure. We met several years ago. You had just started as the chief data officer. You've now built out a global team. We've seen the blueprint that you've created. Customers have begun to adopt that. We've talked to many of them. But give us the update. What's happening here at Think? You've given some talks, and what's new? So, I think you've covered the main points well. It's been about three years. And when I came on board, one of the promises I actually made to our clients was that we were going to make IBM itself into an AI enterprise, and then use those lessons to help our clients make their enterprises, AI enterprises as well. Because a lot of our clients look very much like us. They're large, complicated organizations. So that's the journey we've been on. And we've been progressing on that very well. We created the data and AI backbone for the company. Now we've got various IBM processes that are making use of that backbone to introduce an AI capability, Watson, into their processes. And these range from transactional processes like accounts receivable, all the way to analytical processes like those done by our chief analytics office. The entire platform and backbone is essentially the one that we've built. When we first met, you laid out a prescription of the things that a chief data officer should be focused on. The first thing you said is, you've got to understand the relationship between data and monetization. And a lot of people confused in the early days of big data, oh, I got to monetize my data, I have to package it and sell it. And that's not what you meant. And it could be as simple as, how can you use data to save money? So, how has that gone, that message? How's it gone internally and both externally? Yeah, I think data monetization is all about creating value for the company using data. And there's many paths to it. It depends on the business strategy that the company is following because you want to enable that. That's one way to make money. If they're able to better implement their business strategy because of certain data, then that's going to monetize. And monetize far more rapidly than anything you could package and sell. The other possibility is you could take an operation that's critical to the company and make it a lot more efficient and accurate. That also could release billions of dollars in value. And so it depends on the company itself. So for the case of IBM, another path through monetization is also enabling and helping our product partners, the products that we're using in our data and AI backbone, the IBM products, and we are running through all that so that they can then change their roadmaps based on the actual scaled use cases that we've put together. So there are many different paths to monetization within the company. It depends on the specific case, but eventually it's about tying back to the business strategy and figuring out along the lines of whether you're creating new products, enabling additional revenue or efficiencies or accuracy. It comes back to those kinds of outcomes. So essentially the data value, it's like beauty is in the eye of the beholder. It's contextual to the business. There's no one general purpose data implementation, right? I mean, that's what you're getting at a year. Yes, I mean, it's not so much the implementation, it's the actual path that you take forward. It's got to address certain business outcomes, right? So the generalization is at that level, but one company might pick a very different outcome from another company. And so as a result, what you build, even though the lower levels of the stack might be the same, what you end up delivering and so forth will look very different. Indra, I'll talk about your journey within IBM. I like the narrative of let's do it for ourselves and then share that learnings with the customers. What outcomes were you trying to do internally at IBM to get right and then to bring to the customers? What were those key learnings? What did you learn? What was the outcome? Yeah, no, absolutely. So there are many different outcomes because each process has its own outcome, right? Accounts receivable, they would have like day sales outstanding, that would be a procurement, it would be the time to finish a deal. But eventually you can generalize it by saying it's all around cycle time, end to end cycle time for a process. You want to reduce that and reduce that dramatically using artificial intelligence. So that's been our main outcome that I've been focused on across all our different processes, including my own processes. Now, I think I've mentioned this in the past as well, that eventually it's not so much about technology as it is about other factors that also accompany technology, right? The data itself, how do you prepare it? Make sure that it's ready. But also the cultural aspects of the change, the organizational considerations, the business process changes, people's jobs are changing. How do you make sure that they're trained to do it the new way? How do you tap into the legacy stuff that you've got locked into legacy and then unlock that and make that into AI processes? So there's a lot of work like that that has to go across the lines of not just technology but data, organizational considerations, business process change, and that's the blueprint that Dave was talking about. Jenny made a big deal in her talk yesterday about trust, the stewards of Trump trust, your data. What does that mean specifically from a data standpoint? Does that mean you're not going to appropriate our data to serve ads? Does that mean you're going to secure the data with technology? What does it mean from your perspective? It's again, actually trust cuts all across the stack. So with regard to data and clients, from our standpoint, what that means is the client's data is their data. It's going to remain their data. We will not make use of that data outside of what the client actually authorizes us to do. But not only that, we go even further and we say insights drawn from that data also belong to the client. And the reason we're able to do that is because our business model doesn't depend on the network effect as such, right? In terms of capturing data and then amassing a lot of it, learning from it. You know, getting data from A, but benefiting ourselves and C, right? That's not our model because we're in just in a different world. Our interests are aligned with the client. So it's all about making sure that their data stays their data and the insights also stay their insights. We have no interest in actually capitalizing or monetizing the intellectual capital that our AI systems capture when working with the clients. That's why it's got to remain there. Are those discussions with clients evolving to the point where your commercial terms are evolving? I mean, are they sort of pushing you for different or extended commercial terms that actually explicitly state that and are you involved in that? We, you know, those terms, we just made them available. So clients can pick up those terms if you didn't have to be pushed there, we already knew that this is, you know, because of the nature of AI. And when we started working this within IBM, we realized that AI would become central to every process. Which means that it's capturing not only data, but also the intellectual capital of the company. And then if we put ourselves in the shoes of anybody else, any client who's looking at that, they would want, they'd be very sensitive to how you go about doing that. So we put those terms in right off the bat. So, you know, the clients have, they've got offerings where they can essentially choose. Yeah, this is going to stay, or you can't use it for anything else. Just use it for precisely what we want you to do. That's just part of our standard approach now. I'll talk about this chapter two of the cloud. Jeannie mentioned that. Kind of a nice reference. It's an attention grabber. Okay, chapter two next level cloud. But I want to get your perspective on next level data. What are you seeing the digital 2.0 or the digital generation, the digitization economy happening to processes? You mentioned processes are key. How are processes changing with cloud, with data, with mobile, with these online digital assets and processes? What's changed to these processes that you see generally speaking or specifically? So, one aspect is, and this is why we refer to it as cognitive or augmented intelligence, processes are changing so that the decision makers have access to an intelligent system that helps them do a better job with the decision. Be more accurate, be quicker, et cetera, et cetera. Right? Harness the whole data explosion to our advantage so that you can actually make a better decision. So that's one aspect of the process changes. I think the other aspect is the average enterprise makes use of nine different clouds. So, when we look at that and we begin to understand the complexity that underlies that for an enterprise, right? Being able to manage across these different clouds and when you couple that also with on-prem systems, private clouds, because clients say, well, for our data we really don't want it on a public cloud, we want to do it privately. To manage across all those environments is very tough. It's very difficult. And so, from a data standpoint, you have that same complexity extending into the data space. So now I worry about things like, well, we've got to make sure that if we ingest data once somewhere, we should be able to use it anywhere in an appropriate way, right? In a trusted, governed, secured way. How do you do that? That's an example of the complexities that you have to solve as you go through this new environment. That's the two-dollar. No one you ingested it just to begin with is a good start, right? Yes, but being able to use it everywhere in a way that's secure, I mean, you know, because you're opening up a lot of flexibility, but then you also have to make sure that this is a trustworthy process. So the processes are increasing in terms of capability, decision-making, and efficiency. So you now have more process potential. That's dynamic. Yeah. It's not just that, you know, blocking and tackling straight process. It's baked. We don't touch it. It's getting more dynamic. Yeah, these are, this is new ground, right? Nobody, I mean, that's why I think Jenny drew the distinction between one-dollar and two-dollar. One-dollar was essentially think of it as single-client. Two-dollar is multi-client. And things are different, whether it be from a data standpoint, whether it be from the standpoint of products, you know, now you want products, you run them once, I mean, you write them once, you should be able to run them everywhere. Right, again, appropriately. That's the key part of this, right? In a secure, trusted manner. You can't take something that's running on one side very securely and then you start running it somewhere else and it's no longer secure, right? Then it doesn't work. So, independent of the complexities of hybrid cloud, which you just sort of articulated, what are some of the challenges that you see with regard to people getting their data house in order? I mean, we definitely still see complacency. People say, ah, you know, we're a bank, we're making a lot of money, we don't really have to transform, or, you know, by the time we have to do it, I'll be retired. There seems to be still a sense of, lack of sense of urgency for some customers. What are some of the, is that a challenge, and what are some of the other challenges that you see, even maybe for those guys who want to lean in? I think, at least what I've been seeing over the last three years, that the awareness around AI has increased tremendously. And even within the last three years, you know, the clients now generally don't question that they need to go down that route. They feel the need to go down that route. They begin, they understand that there's a competitive advantage here and there's a danger of being left behind. But their biggest question now is where do I stop? How do I do this in a way that makes me comfortable? Right? So that I don't really end up losing the house while I try to go down that path. And I think that's the central need, that's the central challenge that they face, and that's exactly what we try to do. So they don't want to over rotate to something that's not going to give them a business return. So what do you tell them in that case? Focus on something that's going to drop, you know, save some money to the bottom line, or let's try a little RPA project, or what do you start? You know, what we found is from an AI standpoint, you can do point projects, but you'll only get incremental value by doing those. What you really need to do is to make the whole enterprise, an AI enterprise, so that every process, even, you know, the most, what seemed like the most mundane decisions. I don't know, I mean, I might have told you the story before, Dave, but there's somebody in my organization who labels whether the client we're working with is a government-owned entity or not. That's, yes, and if you think about it, that's, you know, you can think of just this classification task based on what you know, but if you're able to harness the latest news releases, the latest PR releases that are coming out, you're going to make a much better decision, so it becomes an AI task. And think of all the tens of thousands of such decisions that are being made within an enterprise, and you make them more effective through AI. That's the AI enterprise. That's the promise. That's where you're going to get not just incremental change, but monumental change. It'll just completely change the company. Right, so you're saying fundamentally you've got to change the company, and so now there's a cultural aspect of that, which is obviously another challenge. People don't have the skill sets, they don't have the mindset. How are you seeing customers deal with that, and how are you advising them to deal with that? Yeah, so we've been eating our own cooking on this, so we've been through this. We know where the warts are, we know where the pitfalls are, and those are major pitfalls. You have to be prepared to address those. So for instance, retraining the workforce is a major, major aspect that you have to address right off the bat if you go down this path at scale. If you do a point project, yeah, there's no problem, right? You'll make sure you'll be able to do it. Low risk, yeah. But if you're going to do this at scale, then the technology moves very fast. You've got to get the workforce at least comfortable to the extent that they need to do their jobs to be able to use these systems. And so you do need to do that in mass as well, right? Otherwise people will not be able to adopt it, but you won't get the desired return. The point I made about legacy, where literally you could have billions of dollars that are locked in legacy. And so it may not be that easy to apply the AI systems in that context. You have to think through that to get the maximum value of these things. So these are all aspects that go to culture, to change. You know, my boss, he keeps telling me that there are only two words to describe my job. That's not data officer, that's change agent. Yeah, right. Good deal. So we have the wrap. John and I love storytelling. What's the story of IBM Think 2019 from your perspective? Oh, I think it's just been such a dynamic, vibrant conference. I see the energy. I think people are understanding the whole notion of the 2.0 and what it entails as the future is unfolding. And it's just been a terrific conference. Well, it's great to have you on theCUBE again. It's been marvelous to watch your progression over the last three years. Thanks so much for coming on and sharing. It's a pleasure. Thank you both. You're welcome. All right, keep it right there, buddy. John and I were back with our next guest. We're live from IBM Think 2019. You're watching theCUBE right back.