>> Hi, I'm Peter Burris, Chief Research Officer for Wikibon Cube and welcome to another special digital community event. Today we are going to be presenting Wikibon's 2019 trends. Now, I'm here in our Palo Alto Studios in kind of a low tech mode. Precisely, because all our our crews are out at all the big shows bringing you the best of what's going on in the industry, and broadcasting it over The Cube. But that is okay because I've asked each of our Wikibon analysts to use a similar approach to present their insights into what would be the most impactful trends for 2019. Now the way we are going to do this is first we are going to use this video as base of getting our insights out, and then at the end we are going to utilize a crowd chat to give you an opportunity to present your insights back to the community. So, at the end of this video, please stay with us, and share your insights, share your thoughts, your experience, ask your questions about what you think will be the most impactful trends of 2019 and beyond. >> A number of years ago Wikibon predicted that cloud, while dominating computing, would not feature all data moving to the cloud but rather, the cloud experience and cloud services moving to the data. We call that true private cloud computing, and there has, nothing has occurred in the last couple of years that has suggested that we were, anyway, wrong about this prediction. In fact, if we take a look at what's going on with Edge, our expectations that increasingly Edge computing and on Premise technology, or needs, would further accelerate the rate at which cloud experiences end up on Premise, end up at the Edge, and that would be the dominant model for how we think about computing over the course of the next few years. That leads to greater distribution of data. That leads to greater distribution of places where data actually will be used. All under the aegis of cloud computing but not utilizing the centralized public cloud model that so many predicted. >> A prediction we'd like to talk about is how multi-cloud and orchestration of those environments fit together. At Wikibon, we've been looking for many years at how digital businesses are going to leverage cloud, and cloud is not a singular entity, and therefore the outcomes that you are looking for, often require that you use more than one cloud, specially if you are looking at public clouds. We've been seeing the ascendance of Kubernetes as a fundamental foundational piece of enabling this multi-cloud environment. Kubernetes is not the sole thing, and of course, you don't want to overemphasize any specific tool, but you are seeing, driven by the CNC AFT in a broad ecosystem, that Kubernetes is getting into all the platforms, both public and private cloud, and that we predict that by 2020, 90% of multi-cloud enterprise applications will use Kubernetes to lead for the enablement of their multicloud strategies. >> One of the biggest challenges that the industry is going to face over the next few years is how to deal with multi-cloud. We predict, ultimately, that a sizable percentage of the marketplace, as much as 90%, will be taking a multi--cloud approach first to how they conceive, build, and operate their high, strategic value applications that are engaging customers, engaging partners, and driving their businesses forward. However, that creates a pressing need for three new classes of Technology. Technology that provides multi-cloud inter-networking; Technology that provides orchestration services across clouds, and finally Technologies that ensure data protection across multi-cloud. While each of these domains by themselves is relatively small today, we think that over the next decade they will, each, grow into market that are tens of billions if not hundreds of billions of dollars in size. >> The picture I'd like to talk about a very few, the Robotic Process Automation, RPA. So we've observed that there's a widening gap between how many jobs are available world wide and the number of qualified candidates to fill those jobs. RPA, we believe, is going to become a fundamental approach to closing that gap, and really operationalizing artificial intelligence. Executives that we talk to in The Cube; They realize they just can't keep throwing bodies at the problem, so this so called "software robots" are going to become increasingly easy to use. And we think that low code or no code approaches to automation and automating work flows are going to drive the RPA market from its current position, which is around a billion dollars to more than ten X, or ten billion dollars plus by 2023. I predict that in 2019 what we are going to see is more containerization of AI machine learning for deployment to the Edge, throughout the multi-cloud. It's a trend that's been going on for some time. In particular, what we are going to be seeing is a increasing focus on technologies, or projects in code base such as Cube flow, which has been established in this year just gone by to support that approach for containerization of AI out to the edges. In 2019, we are going to see the big guys, like Google, and AWS, and Microsoft, and others in the whole AI space begin to march around the need for a common delatched framework suck such as Cube Flow, because really that is where many of their customers are going. The data scientists and App developers who are building these applications; They want to manage these over Kubernetes using these CNC stacks of tooling and projects to enable a degree of supportability and maintain ability and scalability around containerized intelligent applications. >> My prediction is around the move from linear programming and data models to matrix computing. This is a move that's happening very quicly, indeed, as new types of workload come on. And these workloads include AI, VR, AR, Video Gaming, very much at the edge of things. And ARM is the key provider of these types of computing chips and computing models that are enabling this type of programming to happen. So my prediction is that this type of programming is gonna start very quickly in 2019. It's going to rule very rapidly about two years from now, in 2021, into the enterprise market space, but that the preparation for this type of computing and the movement of work right to the edge, very, very close to the senses, very, very close to where the users are themselves is going to accelerate over the next decade. >> The prediction I'd like to make in 2019 is that the CNCF, as the steward of the growing cloud native stack, they'll expand the range of projects to include the frontier topics, really the frontier paradigms, in micro sources in cloud computing; I'm talking about Serverlus. My prediction is that virtual Kubelets will become an incubating project at CNCF to address the need to provide Serverlus event driven interfaces to containerize orchestrated micro sources. I'd also like to predict that VM and container coexistence will proceed apace in terms of a project such as, specially Kubevirt. I think will become also a CNCF project. And I think it will be adopted fairly widely. And one last prediction, in that vein, is that the recent working group that CNCF has established with Eclipse, around IOT, the internet of things. I think that will come to fruition. There is an Eclipse project called Ditto that uses IOT, and AI, and digital twins, a very interesting way for industrial and other applications. I think that will come under the auspices of CNC in the coming year. >> Security remains vexing to the cloud industry, and the IT industry overall. Historically, it's been about restricting access, largely at the perimeter, and once you provide through the perimeter user would have access to an entire organization's resources, digital resources, whether they be files, or applications, or identities. We think that has to change, largely as a consequence of businesses now being restructured, reorganized, and re-institutionalizing work around data. That what's gonna have to happen is a notion of zero trust security is going to be put in place that is fundamentally tied to the notion of sharing data. So, instead of restriction access at the perimeter, you have to restrict access at the level of data. That is going to have an enormous set of implication overall, for how the computing industry works. But two key technologies are essential to making zero trust security work. One is software to find infrastructure, so that you can make changes to the configuration of your security policies and instances by other software and to, very importantly, high quality analytics that are bringing the network and security functions more closely together and through the shared data are increasing the use of AI, the use of machine learning, etc and ensuring higher quality security models across multiple clouds. It's always great to hear from the Wikibon analysts about what is happening in the industry and what is likely to happen in the industry. But now, let's hear from you, so let's jump into the cloud chat as an opportunity for you to present your ideas, your insights, ask your questions, share your experience. What will be the most important trends and issues in 2019 and beyond, as far as you are concerned. Thank you very much for listening. Now let's cloud chat.

(electronic music) >> Hi, welcome to today's very special in-studio presentation of theCUBE, I'm Peter Burris, Chief Research Officer of Wikibon, and we've got a great guest, we're going to talk about critical infrastructure today, which is a topic that deserves a lot of conversation, but sometimes ends up being a lot of talk and not as much action, and we've got Phil Quade, who's a Chief Information Security Officer of Fortinet to talk about it. Phil, thanks for coming to theCUBE. >> Appreciate being here, thank you. >> So Phil, the issue of security is something, as I said, that's frequently discussed, not often understood, and therefore often is not associated with action, or perhaps as much action as it should be. Talk about the conversation that you're having with customers and peers in the boardroom about the role that security is playing in business thinking today. >> Sure, thank you. The folks I've talked to, they're not dumb people, you don't make it into the C-Suite without having some type of intellect and perspective. What I found is that they recognize indeed that we are in the midst of another computing revolution, and the roots of that trace back from mobility to the cloud and now the Internet of Things. What they don't quite recognize, though, is that we're in the midst of a security revolution as well. And I look at that as going from security from being point solutions to being ubiquitous security everywhere, to having that security integrated so it works as a team. To have that team-oriented security simplified so it doesn't overwhelm the operators. And importantly into the future, much more automation, so highly automated to the degree that it will actually execute the intent of the operator and of the security people. >> So Phil, you made a very interesting point, you said security everywhere, we usually think about security as being something that existed at the perimeter, almost now, I guess, to walking into a building and securing the outside of the building, and once we secure the outside of the building, everything else was fine. But the nature of security everywhere means that the threats seem to be changing. Talk a little about the evolution of some of the threats, and why this notion of security everywhere becomes so important. >> You're right, we all know how well relying on boundary security alone works. It doesn't. You have to have boundary security where there is indeed a defined boundary, but increasingly, networks are borderless. You'll work from home, you'll work from your car. You'll work while you're taking a stroll in the park, but you also need to recognize that you have important assets there in your data centers, there in your clouds, so it's not about having point solutions at the border, it's about having ubiquitous security that can operate in your pocket, on your laptop, on the edge, in the data center, in the cloud as well, but this is importantly, having all those pieces working together as a team. >> We like to talk at Wikibon about the idea of, everybody talks about digital transformation, but to us, that means ultimately is that, companies are using data as an asset, that's the essence of digital transformation. This notion of border security becomes especially important, because our data becomes our representation of us, of our brand, data is acting on our behalf right now. So what are some of those key new things that we're concerned about, in terms of the new viruses? If we think about a hierarchy of concerns, bullying all the way down to strategic, where are we in understanding that hierarchy, and how we're dedicating the right resources to making sense of it? >> Sure, it's tempting to think that WannaCry and NotPetya represent the new normal, or the cutting edge of the cybersecurity threats we're seeing today, but I think we need to take a step back and recognize the intent of such threats. Some threats come at you because someone simply wants to cause mischief. Others because they are trying to bully you into doing certain things. Some of these threats are based on a criminal element, where they're trying to get some type of financial gain, but now others are much more, much more, I'll say harmful. Some might be due to revenge, so, look at the Sony incident. The Sony incident was primarily because a foreign leader was upset of a film company's portrayal of his country, or himself. And the two that are especially worrisome to me are threats that are motivated by military tactical advantages, but most importantly, strategic advantages, so for example, there's some countries that hope to hold our strategic assets at risk, and what I mean is, they'd like to be able to impose their national will on the United States, or other democracies, by holding some of our critical infrastructures at risk, as in preventing their reliable and safe operation, or causing folks to have a distrust of their financial system. So I'm really worried about the threats that come after us from a strategic perspective. Don't worry, WannaCry and NotPetya are important, but they're very different than being strategic threats. >> Now, this issue of strategic threats sounds like there's also a continuum of the characteristics of the threat, from, you totally bring something down, to you actually introduce behaviors that are not expected or not wanted. So talk a little bit about this notion of critical infrastructure, and how we're getting more, both planful, and subtle, and strategic in our responses to the threats against critical infrastructure. >> Well, it's the subtle ones, you're right, it's the subtle ones that worry you, meaning, it's relatively easy to recognize when something bad happens to you, 'cause you can immediately try and fix it, but when something subtle, oftentimes it passes, your prickly sensors don't come up. And the problem is, when all these subtle things build on top of each other, so that all of a sudden, 10 subtle things turn out to be one very big thing, and those are the types of things we need to worry about with some particular critical infrastructures. So for example, a terrorist's malicious activity might simply be looking for one big high-visible attack, meaning, causing heat and light to happen on a TV screen for an exploding oil field, or something like that, but a much more subtle malicious activity would be the gradual degradation of the quality or availability of water, or the gradual degradation on the precision of some of our critical manufacturing, so I'm with you, that some of the subtle things are what we need to worry about. We call those low-and-slow attacks, so it's, you not only be prepared for the loud and stealthy ones, but also the low and slow ones. >> Now, we used to think for example of one of the more famous portrayals of security concerns in movies and whatnot is the idea that I take off the last six decimal places of a transaction, I somehow amass millions of dollars. Is that the kind of thing you mean by low and slow? Those aren't necessarily the kind of threats, I know, but that kind of thing, it's subtle, and it doesn't have an immediate, obvious impact, but over time, it can lead to dramatic changes in how business, or an infrastructure, a national asset, works. >> That's a great analogy, the old financial attacks where they bleed off 0.01 cent per transaction, that adds up very quickly into a very high-volume loss. Well, imagine applying that style of attack on something that could result in not simply a financial loss, but could cause a physical or safety event, whether it be a pressure explosion on a pipeline, a degradation of water, or something of the sort. Those are very, very important, and we need to make sure we're looking for those too. Now, the question might be, well, how do you find such things? And the answer is automation. Human cognition is such that they're not going to be capable of tracking these very low and subtle and slow attacks, so you're going to need to use some always-on analytics to find those types of things. >> So I want to bring you back to a word that you use that, in the context of this conversation, it actually becomes very important. Simple, small word. We. In this world of security, when we start thinking about, for example, the internet, which is a network of networks, some of which are owned by that person, some of which are owned by that corporation, some of which may have more public sponsorship, the idea of we becomes crucially important. We all have to play our role, but to secure critical infrastructure's going to be a public-private effort. So talk a little about how we go about ensuring this degree of control over the public infrastructure. >> So bingo, oftentimes when I say we, it's the royal we, because as you know, as I know, critical infrastructure's not owned and operated by any one place, in fact, it's owned and operated by hundreds if not thousands of different entities. Unfortunately, some people think that the government, the US government, is going to swoop in and do something magical and magnificent to secure critical infrastructure. And the other, certainly, intent, not intent, there's a will to do such a thing, the government doesn't have the authority nor resources nor expertise to do such thing. So what it means is we, this is the royal we, the public sector, the private sector, and there's an even a role for individual citizens, we need to come together in new and innovative ways to get the security critical infrastructure to a much better place. >> And this is part of that conversation, having the conversation about the role that critical infrastructure plays in the economy, in social endeavors, in government, in democracy, becomes a crucial element of this whole thing, so when you think about it, what do the rest of us need to know about critical infrastructure to have these conversations, to be active and competent participants in ensuring that we are having, focusing on the right thing, making the right investment, putting our faith in the right people and corporations? >> I think the first step is taking a long-term approach. I'm a big believer in the old Chinese proverb, a journey of 1,000 miles starts with one small step. The problem with critical infrastructure security is that the problem is so big, and it's so important, that we're often paralyzed into inaction, and that gets back to the point we were talking about earlier, that no one single person is in charge. But we need to recognize that and get past it, we need to recognize that the solution lies in several folks, several communities coming together to try and figure out what we each can bring to this problem. And I believe there's some actional things we can do. I don't know what those 1,000 steps look like to get to where we need to be, but I do know what those first five, 10, 15, 25 things are, as do other folks in the community. So why don't we start acting on them now, and that has the side benefit of not only making incremental progress towards them, but it develops what I call muscle memory between the public and private sector, of how we go about working together on problems where no one entity owns the whole problem, or solution. >> So one of the things that makes critical infrastructure distinct from, again, this goes back to the idea of what do we need to know, is that critical infrastructure is distinct from traditional networking, or traditional infrastructure, in that critical infrastructure usually has a safety component to it, and you and I were talking beforehand about how IT folks like to talk about security, OT folks, or operational technology people, the people who are often responsible for a lot of these critical infrastructure elements, talk about safety. Bring that distinction out a little bit. What does it mean to have a perspective that starts with safety, and figures out how security can make that easier, versus starts with identity, and figures out how to control access to things? >> Right, I think that's an important point, because too often, the folks in the IT, information technology community, and folks in the operational technology community, the OT community, too often were talking past each other, and one of the reasons is just as you said, one focuses on the security of bits and bytes, and other focuses on the safety of water and chemical and electrons and things like that. >> Well, at the end of the day, it's hard to say, "I'm going to secure water by not letting this group drink." >> Right, that's right. >> You can do that kind of thing in the IT world. >> Right. So, very much so, the industrial control system folks, the OT folks, what's number one on their mind is the safety and reliability of their systems and equipment. They're serving the public with reliable transportation, water, electricity, and the like, and so one of the first things we need to do is recognize that, it's not either/or, security or safety, it's both, number one. Number two, I think an important solution is, an important part of the solution is mutual respect, meaning that, yes it's true that the IT folks have some important strategies and technologies to bring into the OT space, but the opposite's also true. The OT folks, some of the smartest folks I know in the business, have been doing what people recently breathlessly call the Internet of Things. So in the critical infrastructure world, they have what's called the Industrial Internet of Things, and they've been using these lightweight distributing appliances for decades successfully. And so I think that we need to take some of the lessons from IT, and apply it to the OT space, but the same is also true. There's some OT lessons learned, so we need to apply the OT space. So, the real solution though is now, taking both of those who are working together to address the increasingly blended critical infrastructures, IT, OT worlds. >> So Phil, if you were to have a recommendation of someone who has worked in, been familiar with the black security world, the black ops world, the black hat world as well as the white hat world, if you were to have a recommendation as to where people should focus their time and attention now, what would it be? What would kind of be the next thing, the next action that would recommend that people take? >> If I could, I'd like to answer that in two parts. First part is, what are the group of activities where we could naturally make some progress? Well, the first one is, getting some like-minded thought leaders together in agreeing that this is in fact a 10-year problem, not a one-year problem. And no matter what jobs we're all in, commit ourselves to working together over that period to get to a good spot, so one is a forming of like-minded people to agree on the vision and determination to help us get there. But then there's some practical things we can do, like, the mundane but important automating information-sharing. There's some critical infrastructures that do that very well today, the financial sector's often brought out as one of the best in that field. But some of the other sectors have a little ways to go, when it comes to automating information-sharing of the threats and the risks in the situations they're seeing. Another thing that I think we can do is, I call 'em pilots. Specifically, we need to explore all the dimensions of risk. Right now when we think about mitigating risk, we think about, how can I stop a threat, or how can I fix a vulnerability. But too often we're not talking about, what are the bad consequences I'm trying to avoid to begin with? And so, the critical infrastructure community especially is ensuring a discipline called consequence-based engineering, so it's mitigating risk by engineering out the bad consequences from the very beginning, and then using your technology to address the threats and the vulnerabilities. So I'd like to see us do some public-private partnerships, some pilots, based on consequence-based engineering, and that will not only reduce overall risk, but it will create, as I mentioned earlier, that muscle memory. >> Consequence-based engineering. >> That's right. >> So is there one particular domain where you have, like when you sit back and say, "I want to see these public-private partnerships," is there a place where you'd like to see that start? Part of the whole critical infrastructure story. >> Right. You can't ignore the electric critical infrastructure. And the good news is that they've been practicing this science, this art, consequence-based engineering, for some time now. So for example, in the electric grid, as you certainly know, there are three major interconnects in the United States, the eastern, western, Texas interconnect. So they already create segments, or islands, so that one failure won't propagate across the whole US. So the mythical US-wide power grid is in fact a myth. But even within those segments, the eastern, the western, and the Texas interconnect, there's other further segmentation. They don't quite call it segmentation, they call it islanding. So when things fail, they fail in a relatively safe way, so islands of power can continue to be generated, transmitted, and distributed. So, in the sense, some of the folks in the electric companies, the electric sectors, are already practicing this discipline. We need to, though, pivot that and use it in some of those other disciplines as well. Think, oil and gas, transportation, water, critical manufacturing, and possibly a couple others. >> So Phil, I find it fascinating, you were talking about the electric grid as a network, and all networks have kind of similar problems, we have to think about them in similar ways, and Fortinet has been at the vanguard of thinking about the relationship between network and security for a long time now. How is your knowledge, how is Fortinet's knowledge of that relationship, going to manifest itself when we start thinking about bringing more networking, more network thinking to critical infrastructure overall? >> You're right, the strategy of segmentation is still king in the security business, and that's especially true in the IT space. At Fortinet, we offer a range of security solutions from the IoT to the cloud, and can segment within each of those different pieces of the network, but more importantly, what we offer is a security fabric that allows you to integrate the security at the edge, at the cloud, in the data center, and other parts of your network, integrate that into a fully-cooperating team of security appliances. What that allows you to do is to integrate your security, automate it much more so, because you don't want to bring a knife to a gun fight, meaning, the adversaries are coming at us in lots of different ways, and you need to be prepared to meet on their terms, if not better. But it also greatly decreases the complexity in managing a network, by leveraging greater automation and greater visibility of your assets. So, you're right. Segmentation is a strategy that's proven the test of time, it's true of the IT space, and it's especially true to the OT space, and at Fortinet, we'd like to see the blending of the planning and implementation of some of these strategies, so we can get these critical infrastructures to a better spot. >> Well, Phil Quade, thank you very much for coming on theCUBE and talking with us about critical infrastructure and the role the network is going to play in ensuring that we have water to drink and we have electricity to turn on our various devices, and watch theCUBE! Philip Quade, CISO of Fortinet, thank you very much. >> My pleasure, thank you. >> And I'm Peter Burris, and I'm, again, Chief Research Officer working on SiliconANGLE, you've been watching theCUBE, thank you very much for being here as part of this very important discussion, and we look forward to seeing you in the future! (electronic music)

(dramatic music) >> Narrator: Live from Fisherman's Wharf in San Francisco. It's theCube. Covering IBM Chief Data Officer Strategy Summit Spring 2017. Brought to you by IBM. >> Hey, welcome back, everybody. Jeff Frank here with theCube. We are wrapping up day one at the IBM CEO Strategy Summit Spring 2017 here at the Fisherman's Wharf Hyatt. A new venue for us, never been here. It's kind of a cool venue. Joined by Peter Burris, Chief Research Officer from Wikibon, and we're excited to have practitioners. We love getting practitioners on. So we're joined by this segment by Bruce Tyler. He's a VP Data Analytics for IBM Global Business Services. Bruce, nice to see you. >> Thank you. >> And he's brought along Fawad Butt, the Chief Data Governance Officer for Kaiser Permanente. Welcome. >> Thank you, thank you. >> So Kaiser Permanente. Regulated industry, health care, a lot of complex medical issues, medical devices, electronic health records, insurance. You are in a data cornucopia, I guess. >> It's data heaven all the way. So as you mentioned, Kaiser is a vertically integrated organization, Kaiser Permanente is. And as such the opportunity for us is the fact that we have access to a tremendous amount of data. So we sell insurance, we run hospitals, medical practices, pharmacies, research labs, you name it. So it's an end to end healthcare system that generates a tremendous amount of dataset. And for us the real opportunity is to be able to figure out all the data we have and the best uses for it. >> I guess I never really thought of it from the vertical stack perspective. I used to think it was just the hospital, but the fact that you have all those layers of the cake, if you will, and can operate within them, trade data within them, and it gives you a lot of kind of classic vertical stack integration. That fits. >> Very much so. And I didn't give you the whole stack. I mean, we're actually building a medical school in Southern California. We have a residency program in addition to everything else we've talked about. But yeah, the vertical stack does provide us access to data and assets related to data that are quite unique. On the one side, it's a great opportunity. On the other side, it has to be all managed and protected and served in the best interest of our patrons and members. >> Jeff: Right, right. And just the whole electronic health records by themselves that people want access to that, they want to take them with. But then there's all kinds of scary regulations around access to that data. >> So the portability, I think what you're talking about is the medical record portability, which is becoming a really new construct in the industry because people want to be able to move from practitioner to practitioner and have that access to records. There are some regulation that provide cover at a national scale but a lot of this also is impacted by the states that you're operating in. So there's a lot of opportunities where I can tell some of the regulation in this space over time and I think that will, then we'll see a lot more adoption in terms of these portability standards which tend to be a little one off right now. >> Right, right. So I guess the obvious question is how the heck do you prioritize? (laughter) You got a lot of things going on. >> You know, I think it's really the standard blocking/tackling sort of situation, right? So one of the things that we've done is taken a look at our holistic dataset end to end and broken it down into pieces. How do you solve this big problem? You solve it by piecing it out a little bit. So what we've done is that we've put our critical dataset into a set of what we call data domains. Patient, member, providers, workers, HR, finance, you name it. And then that gives us the opportunity to not only just say how good is our data holistically but we can also go and say how good is our patient data versus member data versus provider data versus HR data. And then not only just know how good it is but it also gives us the opportunity to sort of say, "Hey, there's no conceivable way we can invest "in all 20 of these areas at any given point." So what's the priority that aligns with business objectives and goals? If you think about corporate strategy in general, it's based on customers and demand and availability and opportunities but now we're adding one more tool set and giving that to our executives. As they're making decisions on investments in longer term, and this isn't just KP, it's happening across industries, is that the data folks are bringing another lens to the table, which is to say what dataset do we want to invest in over the course of the next five years? If you had to choose between 20, what are the three that you prioritize first versus the other. So I think it's another lever, it's another mechanism to prioritize your strategy and your investments associated with that. >> But you're specifically focused on governance. >> Fawad: I am. >> In the health care industry, software for example is governed by a different set of rules as softwares in other areas. Data is governed by a different set of rules than data is governed in most other industries. >> Fawad: Correct. >> Finance has its own set of things and then some others. What does data governance mean at KP? Which is a great company by the way. A Bay Area company. >> Absolutely. >> What does it mean to KP? >> It's a great question, first of all. Every data governance program has to be independent and unique because it should be trying to solve for a set of things that are relevant in that context. For us at KP, there are a few drivers. So first is, as you mentioned, regulation. There's increased regulation. There's increased regulatory scrutiny in pressure. Some things that have happened in financial services over the last eight or ten years are starting to come and trickle in to the healthcare space. So there's that. There's also a changing environment in terms of how, at least from an insurance standpoint, how people acquire health insurance. It used to be that your employer provided a lot of that, those services and those insurances. Now you have private marketplaces where a lot of people are buying their own insurance. And you're going from a B2B construct to a B2C construct in certain ways. And these folks are walking around with their Android phones or their iPhones and they're used to accessing all sorts of information. So that's the customer experience that you to to deliver to them. So there's this digital transformation that's happening that's driving some of the need around governance. The other areas that I think are front and center for us are obviously privacy and security. So we're custodians of a lot of datasets that relate to patients' health information and their personal information. And that's a great responsibility and I think from a governance standpoint that's one of the key drivers that define our focus areas in the governance space. There are other things that are happening. There's obviously our mission within the organization which is to deliver the highest coverage and care at the lowest cost. So there's the ability for us to leverage our data and govern our data in a way which supports those two mission statements, but the bigger challenge in nuts and bolts terms for organizations like ours, which are vertically integrated, is around understanding and taking stock of the entire dataset first. Two, protecting it and making sure that all the defenses are in place. But then three, figuring out the right purposes to use this, to use the data. So data production is great but data consumption is where a lot of the value gets captured. So for us some of the things that data governance facilitates above all is what data gets shared for what purposes and how. Those are things that an organization of our size deliver a tremendous amount of value both on the offensive and the defensive side. >> So in our research we've discovered that there are a lot of big data functions or analytic functions that fail because they started with the idea of setting up the infrastructure, creating a place to put the data. Then they never actually got to the use case or when they did get to the use case they didn't know what to do next. And what a surprise. No returns, lot of costs, boom. >> Yep. >> The companies that tend to start with the use case independently individual technologies actually have a clear path and then the challenge is to accrete knowledge, >> Yes. >> accrete experience and turn it into knowledge. So from a governance standpoint, what role do you play at KP to make sure that people stay focused in use cases, that the lessons you learn about pursuing those use cases then turn to a general business capability in KP. >> I mean, again, I think you hit it right on the head. Data governance, data quality, data management, they're all great words, right? But what do they support in terms of the outcomes? So from our standpoint, we have a tremendous amount of use cases that if we weren't careful, we would sort of be scatterbrained around. You can't solve for everything all at once. So you have to find the first set of key use cases that you were trying to solve for. For us, privacy and security is a big part of that. To be able to, there's a regulatory pressure there so in some cases if you lose a patient record, it may end up costing you $250,000 for a record. So I think it's clear and critical for us to be able to continue to support that function in an outstanding way. The second thing is agility. So for us one of the things that we're trying to do with governance and data management in general, is to increase our agility. If you think about it, a lot of companies go on these transformation journeys. Whether it's transforming HR or trying to transform their finance functions or their business in general, and that requires transforming their systems. A lot of that work, people don't realize, is supported and around data. It's about integrating your old data with the new business processes that you're putting out. And if you don't have that governance or that data management function in place to be able to support that from the beginning or have some maturity in place, a lot of those activities end up costing you a lot more, taking a lot longer, having a lower success rate. So for us delivering value by creating additional agility for a set of activities that as an organization, we have committed to, is one for of core use cases. So we're doing a transformation. We're doing some transformation around HR. That's an area where we're making a lot of investments from a data governance standpoint to be able to support that as well as inpatient care and membership management. >> Great, great lessons. Really good feedback for fellow practitioners. Bruce, I want to get your perspective. You're kind of sitting on the other side of the table. As you look at the experience at Kaiser Permanente, how does this equate with what you're seeing with some of your other customers, is this leading edge or? >> Clearly on point. In fact, we were talking about this before we came up and I'm not saying that you guys led, we led the witness here but really how do you master around the foundational aspects around the data, because at the end of the day it's always about the data. But then how do you start to drive the value out of that and go down that cognitive journey that's going to either increase value onto your insights or improve your business optimization? We've done a healthy business within IBM helping customers go through those transformation processes. I would say five years ago or even three years ago we would start big. Let's solve the data aspect of it. Let's build the foundational management processes around there so that it ensures that level of integrity and trusted data source that you need across an organization like KP because they're massive because of all the different types of business entities that they have. So those transformation initiatives, they delivered but it was more from an IT perspective so the business partners that really need to adopt and are going to get the value out of that were kind of in a waiting game until that came about. So what we're seeing now is looking at things around from a use case-driven approach. Let's start small. So whether you're looking at trying to do something within your call center and looking at how to improve automation and insights in that spec, build a proof of value point around a subset of the data, prove that value, and those things can typically go from 10 to 12 weeks, and once you've demonstrated that, now how do can you scale? But you're doing it under your core foundational aspects around the architecture, how you're going to be able to sustain and maintain and govern the data that you have out there. >> It's a really important lesson all three of you have mentioned now. That old method of let's just get all the infrastructure in place is really not a path to success. You getting hung up, spend a lot of money, people get pissed off and oh by the way, today your competitors are transforming right around you while you're >> Unless they're also putting >> tying your shoes. >> infrastructure. >> Unless they're also >> That's right. (laughter) >> tying their shoes too. >> Build it and they will come sounds great, but in the data space, it's a change management function. One of my favorite lines that I use these days is data management is a team sport. So this isn't about IT, or this isn't just about business, and can you can't call business one monolith. So it's about the various stakeholders and their needs and your ability to satisfy them to the changes you're about to implement. And I think that gets lost a lot of times. It turns into a technical conversation around just capability development versus actually solving and solutioning for that business problem set that are at hand. >> Jeff: Yeah. >> Peter: But you got to do both, right? >> You have to. >> Bruce: Absolutely, yeah. >> Can I ask you, do we have time for another couple of questions? >> Absolutely. >> So really quickly, Fawad, do you have staff? >> Fawad: I do. >> Tell us about the people on your staff, where they came from, what you're looking for. >> So one of the core components of data governance program are stewards, data stewards. So to me, there are multiple dimensions to what stewards, what skills they should have. So for stewards, I'm looking for somebody that has some sort of data background. They would come from design, they would come from architecture, they would come from development. It doesn't really matter as long as they have some understanding. >> As long as you know what a data structure is and how you do data monitoring. >> Absolutely. The second aspect is that they have to have an understanding of what influence means. Be able to influence outcomes, to be able to influence conversations and discussions way above their pay grade, so to be able to punch above your weight so to speak in the influence game. And that's a science. That's a very, very definitive science. >> Yeah, we've heard many times today that politics is an absolute crucial game you have to play. >> It is part of the game and if you're not accounting for it, it's going to hit you in the face when you least expect it. >> Right. >> And the third thing is, I look for people that have some sort of an execution background. So ability to execute. It's great to be able to know data and understand data and go out and influence people and get them to agree with you, but then you have to deliver. So you have to be able to deliver against that. So those are the dimensions I look at typically when I'm looking at talent as it relates particularly to stewardship talent. In terms of where I find it, I try to find it within the organization because if I do find it within the organization, it gives me that organizational understanding and those relationship portfolios that people bring to the table which tend to be part of that influence-building process. I can teach people data, I can teach them some execution, I can't teach them how to do influence management. That just has to-- >> You can't teach them to social network. >> Fawad: (laughing) That's exactly right. >> Are they like are the frustrated individuals that have been seen the data that they're like (screams) this is-- >> They come from a lot of different backgrounds. So I have a steward that is an attorney, is a lawyer. She comes from that background. I have a steward that used to be a data modeler. I have a steward that used to run compliance function within HR. I have a steward that comes from a strong IT background. So it's not one formula. It's a combination of skills and everybody's going to have a different set of strengths and weaknesses and as long as you can balance those out. >> So people who had an operational role, but now are more in an execution setup role. >> Fawad: Yeah, very much so. >> They probably have a common theme, though, across them that they understand the data, they understand the value of it, and they're able to build consensus to make an action. >> Fawad: That's correct. >> That's great. That's perfect close. They understand it and they can influence, and they can get to action. Pretty much sums it up, I think so. All right. >> Bruce: All right thank you. >> Well, thanks a lot, Bruce and Fawad for stopping by. Great story. Love all the commercials on the Warriors, I'm a big fan and watch KNBR. (laughter) But really a cool story and thanks for sharing it and continued success. >> Thank you for the opportunity. >> Absolutely. All right, with Peter Burris, I'm Jeff Frank. You're watching theCube from the IBM Chief Data Officer Strategy Summit Spring 2017 from Fisherman's Wharf, San Francisco. We'll be right back after this short break. Thanks for watching. (electronic music)

(orchestral music) (energetic music) >> Narrator: Live from Fisherman's Wharf in San Francisco. It's the Cube! Covering IBM Chief Data Officer Strategy Summit, Spring 2017. Brought to you by IBM. >> Hey, welcome back everybody! Jeff Frick here with the Cube. I am joined by Peter Burris, the Chief Research Officer at Wikibon. We are in downtown San Francisco at the IBM Chief Data Officer Strategy Summit 2017. It's a lot of practitioners. It's almost 200 CDOs here sharing best practices, learning from the IBM team and we're excited to be here and cover it. It's an ongoing series and this is just one of many of these summits. So, if you are a CDO get involved. But, the most important thing is to not just talk to the IBM folks but to talk to the practitioners. And, we are really excited for our next segment to be joined by Allen Crane. He is the assistant VP from USAA. Welcome! >> Thank you. >> Jeff: And also Glenn Finch. He is the Global Managing Partner Cognitive and Analytics at IBM. Welcome! >> Thank you, thank you both. >> It's kind of like the Serengeti of CDOs here, isn't it? >> It is. It's unbelievable! >> So, the overview Allen to just kind of, you know, this opportunity to come together with a bunch of your peers. What's kind of the vibe? What are you taking away? I know it's still pretty early on but it's a cool little event. It's not a big giant event in Vegas. You know, it's a smaller of an affair. >> That's right. I've been coming to this event for the last three years since they had it and started it when Glenn started this event. And, truly it's probably the best conference I come to every year because it's practitioners. You don't have a lot of different tracks to get lost in. This is really about understanding from your own peers what they are going through. Everything from how are you organizing the organization? What are you focused on? Where are you going? And all the way through talent discussions and where do you source these jobs? >> What is always a big discussion is organizational structure which on one hand side is kind of, you know, who really cares? But is vitally important as to how it is executed, how the strategy gets implemented in the business groups. I wonder if you can tell us a little bit about how it works at USAA, your role specifically and how does a Chief Data Officer eat it, work his way into the business bugs trying to make better decisions. >> Absolutely, we are a 27 billion dollar 95 year old company that focuses on the military and their members and their families. And our members, we offer a full range of financial services. So, you can imagine we've got lots of data offices for all of our different lines of business. Because of that, we have elected to go with what we call a hub and spoke model where we centralize certain functions around governance, standards, core data assets, and we subscribe to those things from a standard standpoint so that we're in the spokes like I am. I run all of the data analytics for all of our channels and how our members interact with USAA. So, we can actually have standards that we can apply in our own area as does the bank, as does the insurance company, as does the investments company. And so, it enables the flexibility of business close to the business data and analytics while you also sort of maintain the governance layer on top of that. >> Well, USAA has been at the vanguard of customer experience for many years now. >> Yes >> And the channel world is now starting to apply some of the lessons learned elsewhere. Are you finding that USAA is teaching channels how to think about customer experience? And if so, what is your job as an individual who's, I presume, expected to get data about customer experience out to channel companies. How is that working? >> Well, it's almost like when you borrow a page back from history and in 1922 when we were founded the organization said service is the foundation of our industry. And, it's the foundation of what we do and how we message to our membership. So, take that forward 95 years and we are finding that with the explosion in digital, in mobile, and how does that interact with the phone call. And, when you get a document in the mail is it clear? Or do you have to call us, because of that? We find that there's a lot of interplay between our channels, that our channels had tended to be owned by different silo leaders that weren't really thinking laterally or horizontally across the experience that the member was facing. Now, the member is already multichannel. We all know this. We are all customers in our own right, getting things in the mail. It's not clear. Or getting things in an e-mail. >> Absolutely. >> Or a mobile notice or SMS text message. And, this is confusing. I need to talk to somebody about this. That type of thing. So, we're here to really make sure that we're providing as direct interaction and direct answers and direct access with our membership to make those as compelling experiences as we possibly can. >> So, how is data making that easier? >> We're bringing the data altogether is the first thing. We've got to be able to make sure that our phone data is in the same place as our digital data, is in the same place as our document data, is in the same place as our mobile data because when you are not able to see that path of how the member got here, you're kind of at a loss of what to fix. And so, what we're finding is the more data that we're stitching together, these are really just an extension of a conversation with the membership. If someone is calling you after being online within just a few minutes you kind of know that that's an extension of the same intent that they had before. >> Right. >> So, what was it upfront and upstream that caused them to call. What couldn't you answer for the member upstream that now required a phone call and possibly a couple of transfers to be able to answer that phone interaction. So, that's how we start with bringing all the data together. >> So, how are you working with other functions within USAA to ensure that the data that the channel organizations to ensure those conversations can persist over time with products and underwriters and others that are actually responsible for putting forward the commitments that are being made. >> Yeah. >> How is that coming together? >> I think, simply put it, it's a pull versus push. So, showing the value that we are providing back to our lines of business. So, for example, the bank line of business president looks to us to help them reduce the number of calls which affects their bottom line. And so, when we can do that and show that we are being more efficient with our member, getting them the right place to the right MSR the first time, that is a very material impact in their bottom line. So, connecting into the things that they care about is the pull factor that we often called, that gets us that seat at the table that says we need this channel analyst to come to me and be my advisor as I'm making these decisions. >> You know what, I was just going to say what Allen is describing is probably what I think is the most complicated piece of data analytics, cognitive, all that stuff. That last mile of getting someone whether it's a push or pull. >> Right. >> Fundamentally, you want somebody to do something different whether it's an end consumer, whether it's a research analyst, whether it's a COO or a CFO, you need to do something that causes them to make a different decision. You know, ten years ago as we were just at the dawn of a lot of this new analytical techniques, everybody was focused on amassing data and new machine learning and all that stuff. Now, quite honestly, a lot of that stuff is present and it's about how do we get someone who adapts something that feels completely wrong. That's probably the hardest. I mean, and I joke with people, but you know that thing when your spouse finds something in you and says something immediately about it. >> No, no. >> That's right. (laughs) That's the first thing and you guys are probably better men than I am. The first I want to do is say "prove them wrong". Right? That's the same thing when an artificial intelligence asset tries to tell a knowledge worker what to do. >> Right, right. >> Right? That's what I think the hardest thing is right now. >> So, is it an accumulative kind of knock down or eventually they kind of get it. Alright, I'll stop resisting. Or, is it a AHA moment where people come at 'cause usually for changing behavior, usually there's a carrot or a stick. Either you got to do it. >> Push or pull. >> And the analogy, right. Or save money versus now really trying to transform and reorganize things in new, innovative ways that A. Change the customer experience, but B. Add new revenue streams and unveil a new business opportunity. >> I think it's finding what's important to that business user and sometimes it's an insight that saves them money. In other cases, it's no one can explain to me what's happening. So, in the case of Call Centers for example, we do a lot of forecasting and routing work, getting the call to the right place at the right time. But often, a business leader may say " I want to change the routing rules". But, the contact center, think of it as a closed environment, and something that changes over here, actually ultimately has an effect over here. And, they may not understand the interplay between if I move more calls this way, well those calls that were going there have to go some place else now, right? So, they may not understand the interplay of these things. So, sometimes the analyst comes in in a time of crisis and sometimes it's that crisis, that sort of shared enemy if you will, the enemy of the situation, that is, not your customer. But, the enemy of the shared situation that sort of bonds people together and you sort of have that brothers in arms kind of moment and you build trust that way. It comes down to trust and it comes down to " you have my best interest in mind". And, sometimes it's repeating the message over and over again. Sometimes, it's story telling. Sometimes, it's having that seat at the table during those times of crisis, but we use all of those tools to help us earn that seat at the table with our business customer. >> So, let me build on something that you said (mumbles) 'Cause it's the trying to get many people in the service experience to change. Not just one. So, the end goal is to have the customer to have a great experience. >> Exactly. >> But, the business executive has to be part of that change. >> Exactly. >> The call center individual has to be part of that change. And, ultimately it's the data that ensures that that process of change or those changes are in fact equally manifest. >> Right. >> You need to be across the entire community that's responsible for making something happen. >> Right. >> Is that kind of where your job comes in. That you are making sure that that experience that's impacted by multiple things, that everybody gets a single version of the truth of the data necessary to act as a unit? >> Yeah, I think data, bringing it all together is the first thing so that people can understand where it's all coming from. We brought together dozens of systems that are the systems of record into a new system of record that we can all share and use as a collective resource. That is a great place to start when everyone is operating of the same fact base, if you will. Other disciplines like process disciplines, things that we call designed for measurability so that we're not just building things and seeing how it works when we roll it out as a release on mobile or a release on .com but truly making sure that we are instrumenting these new processes along the way. So, that we can develop these correlations and causal models for what's helping, what's working and what's not working. >> That's an interesting concept. So, you design the measurability in at the beginning. >> I have to. >> As opposed to kind of after the fact. Obviously, you need to measure-- >> Are you participating in that process? >> Absolutely. We have and my role is mainly more from and educational standpoint of knowing why it's important to do this. But, certainly everyone of our analysts is deeply engaged in project work, more upstream than ever. And now, we're doing more work with our design teams so that data is part of the design process. >> You know, this measurability concept, incredibly important in the consultancy as well. You know, for the longest time all the procurement officers said the best thing you can do to hold consults accountable is a fixed priced, milestone based thing, that program number 32 was it red or green? And if it's green, you'll get paid. If not, I am not paying you. You know, we in the cognitive analytics business have tried to move away from that because if we, if our work is not instrumented the same way as Allen's, if I am not looking at that same KPI, first of all I might have project 32 greener than grass, but that KPI isn't moving, right? Secondly, if I don't know that KPI then I am not going to be able to work across multiple levels in an organization, starting often times at the sea suite to make sure that there is a right sponsorship because often times somebody want to change routing and it seems like a great idea two or three levels below. But, when it gets out of whack when it feels uncomfortable and the sea suite needs to step in, that's when everybody's staring at the same set of KPIs and the same metrics. So, you say "No, no. We are going to go after this". We are willing to take these trade offs to go after this because everybody looks at the KPI and says " Wow. I want that KPI". Everybody always forgets that "Oh wait. To get this I got to give these two things up". And, nobody wants to give anything up to get it, right? It is probably the hardest thing that I work on in big transformational things. >> As a consultant? >> Yeah, as a consultant it's to get everybody aligned around. This is what needle we want to move, not what program we want to deliver. Very hard to get the line of business to define it. It's a great challenge. >> It's interesting because in the keynote they laid out exactly what is cognitive. And the 4 E's, I thought they were interesting. Expert. Expression. It's got to be a white box. It's got to be known. Education and Evolution. Those are not kind of traditional consulting benchmarks. You don't want them to evolve, right? >> Right. >> You want to deliver on what you wrote down in the SOW. >> Exactly. >> It doesn't necessarily have a white box element to it because sometimes a little hocus pocus, so just by its very definition, in cognitive and its evolutionary nature and its learning nature, it's this ongoing evolution of it or the processes. It's not a lock it down. You know, this is what I said I'd deliver. This is what we delivered 'cause you might find new things along the path. >> I think this concept of evolution and one of the things we try to be very careful with when you have a brand and a reputation, like USAA, right? It's impeccable, it's flawless, right? You want to make sure that a cognitive asset is trained appropriately and then allowed to learn appropriate things so it doesn't erode the brand. And, that can happen so quickly. So, if you train a cognitive asset with euphemisms, right? Often times the way we speak. And then, you let it surf the internet to get better at using euphemisms, pretty soon you've got a cognitive asset that's going to start to use slang, use racial slurs, all of those things (laughs) because-- No, I am serious. >> Hell you are. >> That's not good. >> Right, that's not bad so, you know, that's one of the things that Ginni has been really, really careful with us about is to make sure that we have a cognitive manifesto that says we'll start here, we'll stop here. We are not going to go in the Ex Machina territory where full cognition and humans are gone, right? That's not what we're going to do because we need to make sure that IBM is protecting the brand reputation of USAA. >> Human discretion still matters. >> Absolutely. >> It has to. >> Alright. Well, we are out of time. Allen, I wanted to give you the last word kind of what you look forward to 2017. We're already, I can't believe we're all the way through. What are some of your top priorities that you are working on? Some new exciting things that you can share. >> I think one of the things that we are very proud of is our work in the text analytics space and what I mean by that is we're ingesting about two years of speech data from our call center every day. And, we are mining that data for emergent trends. Sometimes you don't know what you don't know and it's those unknown unknowns that gets you. They are the things that creep up in your data and you don't really realize it until they are a big enough issue. And so, this really is helping us understand emerging trends, the emerging trend of millennials, the emerging trend of things like Apple Pay, and it also gives us insight as to how our own MSRs are interacting with our members in a very personal level. So, beyond words and language we're also getting into things like recognizing things like babies crying in the background, to be able to detect things like life events because a lot of your financial needs center around life events. >> Right, right. >> You know, getting a new home, having another child, getting a new car, those types of things. And so, that's really where we're trying to bring the computer more as an assistant to the human, as opposed to trying to replace the human. >> Right. >> But, it is a very exciting space for us and areas that we are actually able to scale about 100 times faster than we were fast before. >> Wow. That's awesome. We look forward to hearing more about that and thanks for taking a few minutes to stop by. Appreciated. >> Peter: Thanks, guys. >> Allen: Thank you. >> Alright. Thank you both. With Peter Burris, I'm Jeff Frick. You're watching the Cube from the IBM Chief Data Officer Strategy Summit, Spring 2017. Thanks for watching. We'll be back after the short break. (upbeat music)

(inspiring piano and string music) >> Announcer: Live from Fisherman's Wharf in San Francisco, it's theCUBE, covering IBM Chief Data Officer Strategy Summit Spring 2017. Brought to you by IBM. >> Hey, welcome back, everybody. Jeff Frick here with theCUBE. We're in downtown San Francisco at the IBM Chief Data Officer Strategy Summit Spring 2017. That's a mouthful, but it's important because there's a series of these strategy summits that are happening not only in the United States, but they're expanding it all over the world, and it's really a chance for practitioners to come together, the chief data officers, to share best practices, really learn from the best, and as we love to do on theCUBE, we get the smartest people we can find, and we have them here. So first off, let me introduce Peter Burris, Chief Research Officer from Wikibon, but from IBM coming right off the keynote-- >> The smart people. >> The smart people, Inderpal Bhandari, he is the IBM Global Chief Data Officer, which is a short title and a big job, and Jesus Mantas, he's the General Manager, Cognitive Transformation, IBM Global Business Services. First off, gentlemen, welcome. >> Thank you. >> Thank you. >> It's really interesting how this chief data officer space has evolved. We've been watching it for years, back to some of the MIT CDOIQ, I think like three or four years ago nobody knew who they were, who were they going to report to, what are they going to do, what's the scope of the job. That's changed dramatically, and it really says something to IBM's credit that they just went out and got one to help really to refine and define for your customers where this is going. So first off, welcome, and let's get into it. How is the role starting to solidify as to what do chief data officers do? >> So, I'll take that. In terms of chief data officers, if you think in terms of the advent of the position, when it started out, I was one of the earliest in 2006, and I've done the job four times, and it has been continuously evolving ever since. When the job was first, in my very first job, I actually had to create the job because there was a company very interested in recruiting me, and they said they sensed that data was critical. It was a company in pharmaceutical insurance, so really very data based, right, everything is driven through data. And so, they had a sense that data was going to be extremely important, extremely relevant, but they didn't really have the position, or they didn't coin the phrase. And I suggested that there were three other chief data officers at that time in the U.S., and so, I became the fourth. At that time, it had to do with, essentially aligning data with strategy, with the strategy of the company, which means how is the company actually planning to monetize itself? Not its data, but itself. And then, essentially make sure that the data is now fit for purpose, to help them with that monetization. And so, that's all about aligning with the corporate strategy, and you have to have an officer who's capable of doing that and has that focus and is able to push that because then, once you start with that strategy, and then, there are plenty of different branches that shoot off, like governance, centralization of data, analytics, data science, and so on and so forth, and then, you have to manage that process. >> And data used to be kind of a liability, hard to think today looking back, 'cause you had to buy servers and storage, and it was expensive, and what do you do with it all? You can't analyze it. Boy, how the world has flipped. Now, data is probably one of your most important assets, but then, the big question, right, what do you do with it to really make it an asset? >> It is, it is, and it's actually fascinating to see here in the summit how even the role that was created in a few years, chief data officer, is coupled with this change in the nature of the value of that role has changed. To your point, I remember meeting some CIO friends 10 years ago that they were telling me how they were deleting data because it was too costly to have it. Now, those same CIOs would give whatever they could have to get that data back and have that history and be able to monetize the data. Because of the evolution of computing, and because now, not only the portion of the physical world that we've been able to represent with data for the last 50 years with information technology, but we're adding to that space all of this 80% of the data that even if digitized we were unable to use in processes, in decision making, in manufacturing. Now we have cognitive technology that can actually use that data, the role of the chief data officer is actually expanding significantly from what used to be the element of data science, of data governance, of data sovereignty, of data security, to now this idea of value creation with basically five times more categories of data, and it actually is a dialogue that we're having here at the summit that is the fascinating from the people who are doing this job every day. >> If you think about the challenges associated with the chief data officer, it's a job that's evolving, but partly one of the reasons why the chief data officer job is evolving is the very concept of the role that data plays in business is evolving, and that's forcing every job in business to evolve. So, the CMO's job's evolving, the CEO's job's evolving, and the CIO's job is evolving. How are you navigating this interesting combination of forces on the role of the CDO as you stake out, this is the value I'm going to bring to the business, even as other jobs start to themselves change and respond to this concept of the value of data? >> People ask me to describe my job, and there are just two words that I use to describe it. It's change agent, and that's exactly how a CDO needs to be, needs to look at their job, and also, actually act on that. Because to your point, it's not just the CDO job is evolving, it's all these other jobs are all evolving simultaneously, and there are times when I'm sitting at the table, it appears that, well, you don't really own anything because everybody else owns all the processes in the business. On the other hand, sometimes you're sitting there, and you're thinking, no, you actually own everything because the data that feeds those processes or comes out of those process is not coming back to you. I think the best way to think about the CDO job is that of a change agent. You are essentially entrusted with creating value from the data, as Jesus said, and then, enabling all the other jobs to change, to take advantage of this. >> 'Cause it's the enablement that that's where you bring the multiplier effect, it's the democratization of the data across the organization, across business roles, across departments is where you're going to get this huge multiplier. >> Yeah, and I think the role of one of the things that we're seeing and the partnership that Inderpal and I have in the way that we do this within IBM, but also, we do it for the rest of our clients is that change agency element of it is the constant infusion of design. Chief data officers were very well-known for the data science elements of it, but part of the constraint is actually no longer the computing capability or the algorithms themselves or the access to the data, which solved those constraints, is now actually preparing the business leaders to consume that and to actually create value, which changes the nature of their job as well, and that's the resistance point where embedding these technologies in the workflows, in a way that they create value in the natural flow of what these jobs actually do is extremely important. Otherwise, I mean, we were having a fascinating discussion before this, even if the data is correct, many business leaders will say, "Well, I don't believe it." And then, if you don't adopt it, you don't get the value. >> You guys are putting together this wonderful community of CDOs, chief data officers, trying to diffuse what the job is, how you go about doing the job. If you're giving advice and counsel to a CEO or board of directors who are interested in trying to apply this role in their business, what should they be looking for? What type of person, what type of background, what type of skills? >> I'll take it, and then, you can. I think it's almost what I would call a new Da Vinci. >> Peter: A new Da Vinci? >> A new Da Vinci is the Renaissance of someone that is, he's got a technology background, because you need to actually understand the mathematical and the data and the technology co-engineering aspect. >> So, if not an IT background, at least a STEM background. >> Exactly, it's a STEM background, but combined with enough knowledge of business architecture. So I call it Da Vinci because if you see the most remarkable paintings and products of Da Vinci was the fusion of mathematics and arts in a way that hadn't been done before. I think the new role of a data science is someone that can be in the boardroom elegantly describing a very sophisticated problem in a very easy to understand manner, but still having the depth of really understanding what's behind it and drawing the line versus what's possible and what's likely to happen. >> I think that's right on. I think the biggest hurdle for a chief data officer is the culture change, and to do that, you actually have to be a Da Vinci, otherwise, you really can't pull that off. >> Peter: You have to be a Da Vinci? >> You have to be a Da Vinci to pull that off. It's not just, you have to appreciate not just the technology, but also the business architecture as well as the fact that people are used to working in certain ways which are now changing on them, and then, there is an aspect of anxiety that goes with it, so you have to be able to understand that, and actually, perhaps even harness that to your advantage as you move forward as opposed to letting that become some kind of a threat or counterproductive mechanism as you move forward. >> I've done a fair amount of research over the years on the relationship between business model, business model design profitability, and this is, there's a lot of different ways of attacking this problem, I'm not going to tell you I have the right answer yet, but one of the things that I discovered when talking to businesses about this is that often it fails when the business fails to, I'm going to use the word secure, but it may not be the right word, secure the ongoing rents or value streams from the intellectual property that they create as part of the strategy. Companies with great business model design also find ways to appropriate that value from what they're doing over an extended period of time, and in digital business, increasingly that's data. That raises this interesting question, what is the relationship between data, value streams over time, ownership, intellectual property? Do you have any insight into that? It's a big question. >> Yeah, no no, I mean, I think we touched on it also in the discussion, both Jesus and I touched on that. We've staked out a very clear ground on that, and when I say we, I mean IBM, the way we are defining that is we are pretty clear that for all the reasons you just outlined, the client's data has to be their data. >> Peter: Has to be? >> Has to be their data. It has to be their insight because otherwise, you run into this notion of, well, whose intellectual property is it, whose expertise is it? Because these systems learn as they go. And so, we're architecting towards offerings that are very clear on that, that we're going to make it possible for a client that, for instance, just wants to keep their data and derive whatever insight they can from that data and not let anybody else derive that insight, and it'll be possible for them to do that. As well as clients where they're actually comfortable setting up a community, and perhaps within an industry-specific setup, they will allow insights that are then shared across that. We think that's extremely important to be really clear about that up front and to be able to architect to support that, in a way that that is going to be welcomed by the business. >> Is that part of the CDO's remit within business to work with legal and work with others to ensure that the rules and mechanisms to sustain management of intellectual property and retain rents out of intellectual property, some call it the monetization process, are in place, are enforced, are sustained? >> That's always been part of the CDO remit, right. I mean, in the sense that even before cognition that was always part of it, that if we were bringing in data or if data was leaving the company that we wanted to make sure that it was being done in the right way. And so, that partnership not just with legal but also with IT, also with the business areas, that we had to put in place, and that's the essence of governance. In the broadest sense, you could think of governance as doing that, as protecting the data asset that the company has. >> They have the derivatives now, though. You're getting stacked derivatives. >> Inderpal: It's much more complicated. >> Of data, and then insight combined, so it's not just that core baseline data anymore. >> And I like to make it an element. You've heard us say for the last five years we believe that data has become the new natural resource for the business. And when you go back to other natural resources, and you see what happened with people that were in charge of them, you can kind of predict a little bit that evolution on the chief data officer role. If you were a landowner in Texas when there was no ability to basically either extract or decline petroleum, you were not preoccupied with how would you protect land rights under the line that you can see. So, as a landowner you have a job, but you were basically focused on what's over the surface. Once actually was known that below the surface there was massive amount of value that could be obtained, suddenly that land ownership expanded in responsibility. You then have to be preoccupied, "Okay, wait a minute, who owns those land rights "to actually get that oil, and who's going to do that?" I think you can project that to the role of the chief data officer. If you don't have a business model that monetizes data, you are not preoccupied to actually figure out how to govern it or how to monetize it or how to put royalties on it, you are just preoccupied with just making sure that the data you have, it was well-maintained and it could be usable. The role's massively expanding to this whole below the line where not only the data is being used for internal purposes, but it's becoming a potential element of a strategy that is new. >> The value proposition, simply stated. >> Jesus: Value proposition, exactly. >> But you're right, so I agree with that, but data as an asset has different characteristics than oil as an asset, or people as an asset. People can effectively be applied to one thing at a time. I mean, we can multitask, but right now, you're having a conversation with us, and so, IBM is not seeing you talk to customers here at the show, for example. Data does not follow the economics of scarcity. >> Jesus: Right. >> It follows a new economics, it's easy to copy, it's easy to share. If it's done right, it's easy to integrate. You can do an enormous number of things with data that you've never been able to do with any other asset ever, and that's one of the reasons why this digital transformation is so interesting and challenging, and fraught with risk, but also potentially rewarding. So, as you think about the CDO role and being the executive in the business that is looking at taking care of an asset, but a special type of asset, how that does change the idea of taking care of the energy or the oil to now doing it a little bit differently because it can be shared, because it can be combined. >> I mean, I think in the way as technology has moved from being a mechanism to provide efficiency to the business to actually being core to defining what the business is, I think every role related to technology is following that theme, so I would say, for example, Inderpal and I, when we're working with clients or on our models, he's not just focused on the data, he's actually forming what is possible for the business to do. What should be the components of the new business architecture? It's this homogenized role, and that's why I kept saying it's like, you could have been one of those Da Vincis. I mean, you get to do it every day, but I don't know if you want to comment on that. >> I think that's exactly right. You are right in the sense that it is a different kind of asset, it has certain characteristics which are different from what you'd find in, say, land or oil or something like a natural resource, but in terms of, and you can create a lot of value at times by holding onto it, or you could create a lot of value by sharing it, and we've seen examples of both metaphors. I think as part of being the CDO, it's being cognizant that there is going to be a lot of change in this role as data is changing, not just in its nature in the sense that now you have a lot more unstructured data, many different forms of data, but also in terms of that's application within the business, and this expansion to changing processes and transforming processes, which was never the case when I first did the job in 2006. It was not about process transformation. It was about a much more classic view of an asset where it's, we create this data warehouse, that becomes the corporate asset, and now, you generate some insights from it, disseminate the insights. Now it's all about actually transforming the business by changing the processes, reimagining what they could be, because the nature of data has changed. >> I have one quick question. >> Last one. >> Very quickly, well, maybe it's not a quick question, so if you could just give me a quick answer. A couple times you both have mentioned the relationship between the CDO and business architecture. Currently, there's a relationship between the CIO and IT architecture, even the CIO and data architecture at a technical level. At IBM, do you actually have staff that does business architecture work? Is there someone, is that a formal, defined set of resources that you have, or should CDOs have access to a group of people who do business architecture? What do you think? >> We've traditionally had business architects at IBM, I think for a long time, that predates me. But again, as Jesus said, their role is also evolving. As it becomes much more about process transformation, it's different than it was before. I mean, this is much more now about a collaborative effort where you essentially sit down in a squad in an agile setting, and you're working together to redesign and reinvent the process that's there. And then, there's business value. It's less about creating large monolithic architectures that span an entire enterprise. It's all about being agile, data-driven, and reacting to the changes that are happening. >> So, turning strategy into action. >> Yes. >> And I think, again, in IBM, one of the things that we have done, our CIO, that is the organization that actually is the custodian of this cognitive enterprise architecture of which Inderpal actually is part of. So, we are actually putting it all together. It used to be an organization. Most COOs have evolved from running operations to defining shared services to now have to figure out what is the digital services version of the enterprise they need to implement, and they can't do that without a CDO in place, they just can't. >> Alright, gentlemen. Unfortunately, we'll have to leave it there. For viewers at home, tune into season two with Inderpal and Jesus. Really a great topic. Congratulations on the event, and we look to forward to the next time. >> Thank you. >> Thank you very much. >> Absolutely. With Peter Burris, I'm Jeff Frick. You're watching theCUBE from the IBM Chief Data Officer Strategy Summit Spring 2017. Be right back with more after this short break. Thanks for watching. (electronic keyboard music)