SDS 137: Cloud Technology: How It’s Changing Data Science, Collaboration and Enterprise

Kirill: Welcome to the SuperDataScience podcast. My name is Kirill Eremenko, data science coach and lifestyle entrepreneur, and each week we bring inspiring people and ideas to help you build your successful career in data science. Thanks for being here today, and now let’s make the complex simple.

Kirill: Welcome back to the SuperDataScience podcast, ladies and gentlemen. Super excited to have you on board. Today we’ve got a very interesting guest, Derek Schoettle, who is the general manager and chief business officer for the Watson and cloud platform at IBM. It was very exciting for me to host Derek because he’s made quite a few appearances at different conferences. If you look for YouTube videos with him, you’ll find him talking about the cloud and what this means for business and how businesses can transition to the cloud and what kind of platforms and offerings IBM has in this space, and that’s exactly what we spoke about today. This episode is entirely dedicated to the cloud and what it means and what it could mean for your business or what it can mean for your career. So if you have no idea whatsoever about the cloud, then not a worry, Derek will explain all of it, and if you’re really well-versed in the space of cloud, then you will learn about where the world is going in terms of cloud computing, in terms of cloud platforms and what that means for data science and business.

Kirill: Also, we took a little bit of a detour towards the end and we spoke about executives in the space of data science, and throughout the podcast you will get hints, if you’re an executive or if you are a manager or if you’re a business owner or entrepreneur, throughout the podcast you’ll get hints on how to build a data-driven business. Specifically at the end we talked about how executives can upskill in the space of data science in this day and age. Very, very interesting podcast all and all, and can’t wait for you to check it out. So let’s dive straight into it. Without further ado, I bring to you Derek Schoettle, the chief business officer of the Watson and cloud platform at IBM.

Kirill: Welcome, ladies and gentlemen, to the SuperDataScience podcast. Today I’ve got a very special guest on the line, the chief business officer for the IBM cloud platform and IBM Watson, Derek Schoettle. Derek, welcome to the show. How are you doing today?

Derek: Good. Thanks for having me. Excited to be here.

Kirill: We were just talking before the podcast about the weather in Boston. Is it always like that?

Derek: No, I think most people in New England live here for the bit of weather we get from kind of May until November, and then from November until May we kind of just grin and bear it.

Kirill: Yeah. When I was in Boston I didn’t really actually notice the Boston accent. Is that a thing? In macho movies and stuff like that?

Derek: Yeah. Yeah. So I was I was born in Minnesota. I was then raised in Manhattan and New York City, and I split time between the two. I never thought, given if you grow up in New York City, Boston is a place that you think of as kind of a kid sister or a little cousin. But then I end up going to school up here. I ended up starting a family here. The one piece that you come to love is the Boston accent is very real, and at first it’s not something that you’d necessarily enjoy hearing all the time, but it’s strangely endearing. It’s like a Bronx accent in New York, and I’m not going to even try and attempt to mimic it. I don’t have it. I don’t have that that charming regional gift yet.

Kirill: Interesting, okay. So you think it’ll come with time?

Derek: Well, I’ll tell you, my wife and I have four kids. My son’s 13, I have three daughters who are 12, 10 and 9, or 12, 10 and 8. I suspect one of them will be able to answer that call, if not all of them at some point perhaps.

Kirill: Okay, fantastic. All right, well, today thank you very much for coming on the show. It’s great to have you. Today we’re talking about cloud and Watson, the IBM Watson the legendary program that beat humans at Jeopardy and is doing amazing things in the cloud, but more importantly about analytics in the cloud and data science in the cloud. So you are the general manager or the chief business officer at the Watson and cloud platforms at IBM. Can you tell us a little bit about your role? What does the day-to-day of the chief business officer look like?

Derek: Sure. I mean, at the end of the day, my success or our collective success is, and my role is focused on delivering a platform. So responsible for taking what we build, how we build, how we bring it to market and then most importantly, how it’s positioned within the broader market as it relates to developers, partners, the broad IBM community of partners, and that day-to-day looks like working on centralized operations, working on offering management, product management, working on go-to-market campaigns, basically the full operations of the business. Now, where I enjoy spending most of my time if given the option is around product development and go-to-market.

Derek: I think, coming from the companies that I’d spent time at prior to IBM, they were smaller technology companies that had in cases reached scale, but there is nothing like working in a place like IBM where there’s such a rich heritage, there’s immense talent, there’s never-ending, I’ll say, pool of technology, and really, what I think in this day and age, this platform economy, my aim is to kind of make what is complex simple and what is hard to understand and reach accessible. And if we can deliver on that, I think IBM and our clients and kind of, again, that broader ecosystem, really do stand to benefit.

Derek: It is an amazing time to be in technology in a day and age when you now have compute, you have network, you have storage, you have massive amounts of data and you now have a pretty fast evolving AI landscape. And those ingredients because you’ve seen different fits and starts with machine learning, with artificial intelligence technologies, you’ve seen fits and starts with data and analytics, you’ve seen fits and starts with managed services, or now what we’re thinking of as the cloud, but I don’t think it’s until the last couple of years that they’ve all come together in a way that the technology, the economics and most importantly, the business value, the value for the end-user, for the clients that we work with can really reap the rewards. So it’s an exciting time to be here. So I view my role for IBM as deliver a platform, align the broad set of resources we have, all towards a path so what we provide our clients is, again, simple and accessible to consume and use and scale.

Kirill: Totally. I agree with you that this is probably one of the most exciting times to be alive right now. In the space of cloud what I’ve personally seen is that about like five or so years ago people were very still skeptical and afraid of security issues and so on, and companies were hesitant to move into the cloud. What are you seeing now? Are you noticing any larger uplift in the uptake of the cloud services?

Derek: Yeah, it’s a good question. I can recall in 2005 I was at a messaging company where we were working in the early stages of what is now AWS, and the thought of trusting a third party with your email was almost sacrilege. But myself and my business partner at the time kind of, like many I think, saw that this was going to be a trend that would sustain. Just given Moore’s law, right? I mean, based on where compute was going and based on where the overall economics were heading.

Derek: I do think that the broader market has awoken to the fact that a cloud utility is very valuable. I think what we continue to see, and I think where you’ll see IBM spend a lot of its time and energy and investment is this very real Enterprise perspective that, well, enterprises do realize the value of the cloud. I think there is an existing estate, there’s an existing business that’s being run on systems and equity, right, skills that have been developed that are not necessarily just to be jettisoned but rather migrated to, brought forth into the public cloud at a rate and pace that makes sense for their business and their clients and I’ll say their operating environment, what relative risk and compliance and security is expected of them in order to be relevant and competitive.

Derek: So from our standpoint, yes, we see a lot of interest in and we see it globally, right? I think the first population to really move here starting five years ago was in predominantly North America, and then it started going to Asia and to Europe and India, obviously, and globally. I think in the last couple of years we’ve seen that become the norm versus the exception. But make no mistake. The enterprise expects a level of security and a level of flexibility in terms of how they deploy their offerings and their products. It’s very much IBM’s perspective that we are solving an enterprise cloud problem. We are not the cloud for everyone. We are really the cloud for enterprise.

Kirill: Got you. Could you for those listeners out there who might not be completely up to date with what cloud means these days, could you in layman terms explain, like in a couple of sentences, what is the cloud?

Derek: I’ll do my best. How I think everyone should think about the cloud is it is a chance to, as an individual or a customer, take advantage of compute, network, storage, software at a rate and pace, meaning what you pay and what you consume are directly linked. So it’s a utility. And you can consume it at a rate that’s appropriate for your business. So if you’re just wanting to explore a technology, or if you’re just wanting to prototype an application, you’re not forced into or you’re not kind of required to invest huge sums of money and incur, I’ll say, a lot of different interfaces to get to that value. The cloud is really designed to provide immediate access to the right technology at the right time at the appropriate kind of economic structure or pricing so that value and offerings are very closely linked.

Derek: I’d say the other dimension to think about is it’s really an endpoint. I think the traditional interpretation is the cloud is something that your grandmother thinks is a drawing, and it’s just up in the blue sky, and no one really knows what it is. I think the more modern realization is that it’s just infrastructure. It’s an endpoint for certain applications, certain workloads. And I think as we make our way into the, I will say, after this kind of the existing track we are on with cloud, what is next with serverless and edge computing and the emergence of IOT platforms, you’re going to continue to see evolve where you want to be able to build and deploy your applications in the right infrastructure at the right time with the right terms. I think that speaks to what I was touching upon just a moment ago that our view with IBM is you don’t want to have to make a choice that is specific to an underlying deployment. You want to be able to make a choice that’s appropriate for your data, your existing technologies, where you want to invest net new. That can be on-premise, that can be hybrid, that can be in public cloud. Our enterprise cloud meets all of those demands. I’ll say last comment is that the cloud is I think really helped business.

Derek: Think of computing and network and storage and software on a more consumer bent, meaning self-service, discover, try and buy, remaining anonymous as you go through that discovery process, where 10 years ago, 15 years ago if you wanted to purchase a database or you wanted to purchase storage or get onto any kind of platform, more often than not you’d have to engage in a series of discussions, you’d have to read white papers. You’d try and find a pure set where they could get advice or perspective. You’d talk to a handful of analysts. Whereas I think what the cloud has helped us is it’s really changed the buying motion. It’s changed the way that people look at and try technology in a way that is very specific and anonymous. I like to think that, and I think this is a good thing. I think it’s good for all of us is that it’s allowed people to make decisions with far more accuracy and immediacy because you can see the code, you can see the product, you can see what the actual offering is far more so than you’ve ever been able to in the past. So we very much see this kind of self-service world that is allowing a lot of groups to come together in a common platform, working together at a rate and pace that’s appropriate to them. I think the cloud enables all of them.

Kirill: Well, fantastic. Very, very detailed description, thank you so much for that. I was once at a company that was in front of a dilemma whether to upgrade their servers because they outgrew their current capacity and performance wasn’t the best, they were either going to upgrade their servers or go into the cloud and upgrading the servers was $20 million, just that they had to spend right off there, or they would go into the cloud. I’m still not sure what exactly they did because I’m not working with them anymore. But in terms … That’s one advantage of the cloud, right? It’s very easy to scale. You just click a button, you get more. What are some of the other advantages of the cloud that you can share with our listeners?

Derek: Yeah. There’s the scale piece which is you benefit from the scale at which that underlying platform operates which shows up in durability. It shows up in the economics, meaning the pricing because you’re getting the benefit of that scale. I think the other thing is choice. I think you’re given a chance to work with a lot of different technologies, where probably the example you’re just giving about the company you worked at before, once you make a considerable capital investment in any technology, you’re somewhat limited to that investment, whereas again, on a utility platform where you’re given a high degree of choice but with consistent delivery, that’s a huge advantage. So from a developer standpoint you have a range of tools, you have a variety of underlying infrastructure and performance attributes that you can use, and you have a whole set of resources at your disposal, and be that communities, be that documentation, forums. I just think the immediacy that the cloud provides is really compelling.

Derek: I think the other thing is that … I’ve done earlier stage technology companies in my career. The first part was, well, if I’ve got customers in Europe or Asia, or if I’ve got customers that have unique demands that I alone with my own data center, with my own capabilities would really be hard-pressed to be able to fulfill, today we can meet customers’ needs anywhere and can do so with the push of a button and still maintain the integrity and the quality of service that we’d want to be able to provide. I think that, in terms of Tom Friedman’s book, the world is flat, the clouds help flatten everything. We can deliver goods we can provide access to and the value of data at a scale that you couldn’t before without the cloud. I don’t think that can be overlooked, and I think it’s just the beginning. I think this is chapter one of multi trilogy and many chapters. It’s very early.

Kirill: Totally agree. You mentioned that a big part of your job is delivering the platform, and what I want to talk about here is that IBM didn’t just stop at providing this cloud service where you can go and you get that scalability and cost-effectiveness, but you guys are actually working on, you have released, have launched a platform called The Data Science Experience. I’d like to talk a little bit more that because I think that will be very exciting to our listeners on this podcast.

Derek: Yeah, for sure. What I’m most excited about is, and this is how we kind of think of the world is, in this new data economy, you really do need to think of where you’re putting your data and does it sit within a platform that’s been designed for it so that the underlying governance and management and accessibility and economics of the data itself is at the forefront, right? So that’s the kind of cloud platform that we have. It’s called the Watson data platform. That platform is what we’ve now launched a suite of applications on top of for data scientists. The really interesting dimension here with Data Science Experience is at its core a notebook, and then a set of tools around that notebook, like you can code in Python, Scala. We have something called PixieDust, which is a simple way to kind of deprecate some of the coding challenges that you can experience, so that the every person can interact with a notebook versus a select few. It’s a simple iterative way to start designing and training and deploying machine learning models, analytic assets.

Derek: I think there’s kind of the … think of that notebook. Best-in-class set of tools. What’s really cool is we’ve taken this dimension where it is collaborative. We very much think of data science as a team sport, right? It is not one person locked away with a model and a training set and some expectation of having to come up with a recommendation, they flip over the wall to either the marketing organization or the trading desk or IT operations. The Data Science Experience running on our platform is meant to be a collaborative environment where your business analyst, your data scientist, your developers, your marketing professionals, all the people that are in and around this intersection of … I was a retailer. How do we build the best interactions and position our products in the most efficient way? Well, that takes a lot of different perspectives all working on and with data. So this offering is meant to help people come together in a collaborative environment that allows you to move quickly, find data easily and catalog that data.

Derek: Let’s say, once you’ve created a machine learning model or you’ve got a Spark job or you have a notebook where you have a dashboard that you’ve used and you’ve enjoyed success, all of those become assets on this platform that we catalog for you. Think of what Spotify has done for music. What we want to do is with our data catalog is do the same thing but for enterprise data, right? So if you remember what I was talking about with self-service analytics and being a provide access to the right data at the right time for the right teams, Data Science Experience brings all of this together. It’s self-service. It’s free to trial. You can onboard your data, start building analytic assets, training models.

Derek: It’s one of these things that we launched a little over a year ago. We now have tens of thousands of data scientists running on the platform today. We’re working with a really interesting cross-section of retailers, financial services companies, oil and gas, shipping and logistics, and it’s beginning to grow virally where teams are talking to teams are talking to teams, and the common kind of plane or the common network is data. That I think is really cool. I think this is one of these things where if you look back at how business intelligence evolved in the early 2000s, where you had traditional client-server and all of the data was taken out of the database, put into this thick client, cubes were created and it was kind of small data, and you’d manipulate the data as best you could, and then you’d send out a report once a week, to now in the world that we’re in today. It is team oriented, it’s meant to be iterative in real time, and it’s meant to be an experience that takes you from iteration to production as quickly as possible. Not weekly reports or whole days pass as you try to evaluate different options, it’s meant to be at the speed of business which is, I think, where this is all heading, right?

Kirill: Yeah. I totally feel that, a pane of the cubes and the reports once a week. I’ve been there, done that, that’s not fun at all.

Derek: That is not fun. I think it’s really interesting companies, us included, that are really changing the dynamics so that one part of the equation is how do you find the data, and we offer a set of tools as a part of Data Science Experience to find the data, to refine it, to put it in a package that allows you to then interrogate it, right? To work with it. Simple dashboarding and visualization, embedded URLs, embedded kind of runtimes so that we really bring data to life for teams. It’s what we’re trying to do.

Kirill: Fantastic. I’m very impressed that you guys went with the open source tools like IronPython. I think that’s a big advantage for lots of people out there. These are growing tools that are taking a huge share of the market. And still, some proprietary tools will or some platforms will not incorporate IronPython. I think that’s a mistake. So what is the decision there?

Derek: You have to … Yeah. One of our key themes is this is an open platform, right? We want to reach across different systems, different environments regardless of where the data may reside so that we don’t force people that are comfortable with their weapons, their tools of the trade, their resources, we don’t force that. We rather say, come into our environment, right? And the notebook is the centerpiece, and we want to give you the best tools so that you trust and you can work with the right data, and then you’ll use our platform to put it into production, that full lifecycle. I think the decision to, I’ll say, strict construction of those tools is not something that, based on all of our experience over the years of … I mean, we have a very large information management business, SPSS is one of many products that are in and around the space that we’ve learned from, and I think what we learned there and continue to see from our clients is they want choice but with consistent deployment.

Kirill: Got you. Another thing I wanted to ask, I think this might be relevant to this Data Science Experience is the notion that you guys have highlighted, that you are able to help enterprises resolve the 80/20 dilemma, where right now data science teams are spending 80% of their time on data prep. Personally I have had the same experience. So it’s a very bold statement. I’m very interested in how you are going about that.

Derek: Yeah. So we have a set of tools called Data Refinery. The thing we say is if you’re going to chop down a tree, spend 90% of the time sharpening the ax, right? I think, historically, the data challenge has been that you have data quality issues, you have data access and you have data predictability, all of those we’d hope to solve in our refinery where we put the necessary tools at the hands of the teams to say, go find your targets, find your source data, apply them to different targets, meaning put it into an object store, catalog it within data catalog and then expose it within a notebook for analytics, that refinery exercise is the first thing you do. Within Data Refinery you’re able to discover the data, you’re able to drop columns, you’re able to drop nulls, you do basic data cleansing. We have algorithms that will crawl the data to give you out-of-the-box data quality scores so that IT sets up this platform with a set of identified users, those users then have rights, and those rights have then access to different data sources.

Derek: Those data sources are then positioned … And you can think of an Oracle database, Redshift, S3 Salesforce, pick any of your kind of semi-structured or structured data sets that we want to be able to provide our users to find, refine and then put into our data catalog to then put it into our Data Science Experience. That is where we’ve spent a lot of time. I think the trick there is having the right number of connectors to those largest kind of native sources and then being able to easily bring that data to, whether it’s metadata that you’re bringing, right, a metadata catalog, or it’s a subset of the total that you want to be able to train models on, or that you want to build analytic assets with.

Kirill: Okay, I can see how that works. You’re not just automating the machine learning part of things, but also the data preparation. I think that’s an important step.

Derek: Yeah, the discovery part. I mean, that to us is this whole team concept. So what are the tools the teams need in order to find, onboard and prepare data for a variety of different analytic tasks? That task can be the iterative practice of refining a notebook, it can be creating dashboards that you share or that you post to websites that update automatically as new data flows through the system, or it’s deploying models at the service of, whether it’s an e-commerce site or it’s a trading application, whatever the end application may be that’s ensuing the model, it’s that full lifecycle. That goes back to for us, the data platform you can think of as an operating system, and this suite of applications that run on top of it as kind of our suite for data and AI.

Kirill: Okay. Got you. What about AI? How’s that going? Are more and more clients starting to pick up AI?

Derek: Yeah. What I would say here is that, first, you think of Watson as a brand, and we have for this next era of learning that’s going to be accelerated through different technologies that we would put under the banner of AI, and that can be machine learning, it can be deep learning, it could be natural language understanding, it can be inference, speak to text, all of these different, I’ll say, capabilities, that when brought together allow, whether it’s a customer agent, whether it’s a wealth management professional, whether it’s an application, whether it’s two bots working together, all of them can be harnessing that underlying substrate of AI technologies.

Derek: So our platform and the Data Science Experience, and this is going to be coming in the next couple of months, we’re gradually disclosing more and more of the underlying AI technologies above and beyond machine learning and deep learning. You can imagine a team, right? So that, let’s say, we go back to that retail e-commerce site where they need to be able to come up with better offers and offers that are delivered in real time, and they want to be able to do that through a virtual agent. So exposing the virtual agent technology into the development cycle. So as you’re sitting within that IDE, i.e. the notebook, and you’re loading data, you’re training models, and you’re iterating around how you then embed that within a virtual agent that it’s embedded within your website. Today that is seven to eight different steps that you need to take across multiple different technologies.

Derek: Our platform, back to this operating system metaphor, our data operating system is what you then build these applications on top of. So those capabilities, machine learning, deep learning and the suite of AI tools embedded within Data Science Experience, embedded within that development environment. So marketing and sales and development and data science are all able to interact with one another in a collaborative model so that the data gets to the right spot, the applications that are built, again, with this retail example are presented in the right way. And as that campaign goes live or as that virtual agent is launched and you’re watching it and you’re making sure that it’s delivering what we’d expect for the customer experience and it’s moving more product and improving, improving, improving, the whole team is engaged in that. It’s not off in silos. The whole team is working in this application suite sitting on the platform.

Kirill: Okay. Okay, got you. While we’re on this topic of AI, I just wanted to touch on cognitive science and cognitive computing because that’s what I’ve heard of IBM Watson. Could you just demystify that term for us quickly and then we’ll move on from there?

Derek: Yeah, cognitive is the intersection of what a machine is able to do to mimic the human brain, right? That is the set of technologies and investments that we’ve made under the Watson brand. The manifestation of that can be in a physical appliance so that the chipset, the underlying compute, the appliance itself is optimized for AI. So if you look at our POWER9 chip that’s designed for AI, that’s an example of a cognitive system, right, because it’s been designed for and very much positioned as something where you should train and deploy and manage your AI platform.

Kirill: Got you, thank you. Speaking of data science as a team, there’s this interesting thing that I think you guys are doing quite well is the developer/data scientist relationship and how the tools that you create can bring them together. What are your views on how these two absolutely different areas, development and data science are coming closer together and why that’s happening and what’s the future behind that?

Derek: Yeah. It’s a great question, and one that we spend a lot of time talking to our clients about, and we see in the broader kind of marketplace is that more and more, right, if you look at the development lifecycle, I would say agile is more the norm than the exception, right, where years ago you’d have waterfall development where you’d release a product every quarter and then dot releases, or maybe a major release every six months and a dot release every quarter. I think now that we’re in the cloud economy … I can’t believe I didn’t say this earlier, huge benefit of cloud also is the rate of innovation because you’re updating these technologies five times a day, many times a week, all at the benefit of, say, one customer wants an attribute or a feature. As soon as the feature is updated, everyone benefits, right? There’s an accelerant that happens, kind of a network effect. So the same premise of how and why teams are operating differently as a function of that rate and pace, right?

Derek: So products are getting developed far more quickly. Prototypes are getting put into production, sample applications are getting put into production environments and live environments so that clients can start interacting with them. What is really important about that is the development team launches an application, the speed at which they then figure out how is the application performing, how does it need to change, what additional features need to be added, what data is this application creating and how can that help influence not just those users of that application but perhaps adjacent or other capabilities that you as a provider may want to surface, that lifecycle, that operational lifecycle is what see blending, right? Where historically it was development flips it over the wall to QA, QA flips over the wall to IT operations, IT operations then puts it into kind of runtime, and then there’s a whole set of different people in a constellation observing how that application is behaving.

Derek: I think the world that we live in today and it’s only becoming more so is that teams are looking at applications in real time and evolving them in real time, and what allows that collaboration and that teamwork to happen is our platform and the tool chain that we’re providing them and the operational runtimes that we give them so that you shorten the cycles between initial discovery and action. The hope is that you’re using the best data at the right time for the right decision as a team so that your decisions are made with all of the appropriate inputs versus just one perspective, you’re getting the benefit of the entire group.

Derek: I look at Slack as a great example of this where you have a real-time collaborations platform that has stormed into the enterprise and stormed into a team dynamic far more quickly than anything I’ve ever seen in my life, and I think because they are tapping into this reality that all of us are needing to move so quickly, we’re all needing to move and these squads and teams, and we’re needing to share information, operate with a higher degree of transparency and communication, on and on and on and on and on, loosely coupled, closely aligned team of teams, that’s happening for data scientists and developers and business analysts and operators. All of them are coming together. And I think in order to do that, you’ve got to have them run on a platform and a consistent application suite that allows them to do that.

Kirill: Yeah. I totally agree with that. I think while we’re witnessing a more of an [inaudible 00:35:52] socialization in real life where people are talking less, meeting up less and so on, is becoming more collaborative online and tools that leverage that, that team mentality, they’re going to be the ones that win.

Derek: I mean, look at office spaces. If you go to an office at perhaps an older organization, like IBM, there’s office buildings where long hallways, every office has a door and you’re sequestered into your office and the door closes, that’s now the exception, not the rule. Now, everyone’s in open, free-forming workplaces where walls are moveable, you can write on everything, teams can form naturally and then kind of separate and come together in different settings. Technology is no different, right? Our platform and our applications want to model that kind of work environment. How people want to work? They want to work in teams, agile-y, with their choice of tools, on a trusted and secure environment. That’s our whole premise. Data science is a team sport and our platform aims to make data simple and accessible for the world. And if we can deliver on those two, as we have been and hope to continue, I think we have a chance to really help a lot of clients with problems that they’ve been struggling with today because collaboration is not as native, because data is not as is accessible and available as it needs to be.

Kirill: Okay. Thank you very much. A great overview of that. Now shifting gears a little bit, I’ve got a quote here from your blog. You said that data has become the most valuable currency and the common thread that binds every function in today’s enterprise. The more an organization puts data to work, the better the outcome. My question would be there would be about the strategy. Data strategy, data governance, a culture that supports data and data science. What are your views on that, and what are your recommendations to listeners who might be executives, directors, entrepreneurs, business owners, how to think about data strategy?

Derek: Yeah. My favorite quote on this one is, if you don’t have any data, we’ll just use my opinion. It’s meant to be a challenge to anyone who’s in an executive leadership position is that in this day and age, where the surface area of what all of us are being asked to understand and make decisions about, we don’t have months to slowly think and process and make decisions. I mean, obviously, there are those decisions you have to take that time to think through, but in many cases, you have to partner with data. You have to make data present in all of your discussions and almost have a seat at the table so that, as you’re moving quickly, as you’re working in this new team dynamics, data helps ground you in the truth. One great example for us is we’ve put Net Promoter Score at the forefront of how we measure our success as teams, delivering technology, providing support, engaging with clients, providing services to them across the full spectrum of all that we do within our business. That data is a really grounding and healthy contribution to everything we do because what it does is it forces us to be honest about what’s working, what’s not working, and it puts our customers’ voice right at the epicenter of what we’re doing.

Derek: I think that’s but one of many examples where in today’s day and age where we have the technology, we have the volume, the velocity, the variety of data that we can now use it to inform our everyday lives, you should embrace that, not run from it. I think it’s a cultural change, right? I mean, a lot of times, I think, in our collective past as a society, as an economy, people were able to make decisions through influence and make decisions through perhaps, bending, quote-unquote, the truth. Data doesn’t necessarily lie, right, especially if you set it up on a platform that’s trusted, it’s governed, it’s secure, and you can use that as a backbone for how you run your business. It invites people into the conversation, right? If everyone can join the conversation around a common dataset, you get a level playing field, you get more participation, you get more equality.

Derek: I think we drive in our group a very flat organization. We try, and whether you want to call it long leashing, loosely coupled, closely aligned, we try and run very flat so that people are able to move the pace that our clients want us to move at. I think data allows us to do that because it keeps everyone kind of aligned on, okay, what’s most important for us? How are we sure that we’re delivering the right value for our clients at the right time?

Kirill: All right. But that’s a very powerful way for you guys to go about it, and IBM has been in this space forever, and you have developed a strong culture in that sense, but what about a business owner who’s just starting out, or who, I don’t know, like an executive who just took over at this executive position, and they see that the data culture is not there and the organization will fail in the next five to 10 years unless they fix that. What is your advice there? How do they go about embedding that from scratch?

Derek: Yeah, it’s a good question. I think what I’ve done in every leadership position before IBM, so whether it was at Cloudant, whether it was at Vertica, Intellireach is, at first, you set up your kind of the ways in which you’re going to operate, the operating principles of the business. We have them within my team at IBM, and it’s over communication, it’s ultimate transparency, it is ultimate accountability, meaning, this is our business, it’s ours to run, it’s ours to … failure is going to happen, embrace it. Things of that nature so that one you set up the operating environment. And then, two, you find corresponding metrics that data can help fuel, right? So decisions about how you’re performing. Everyone should agree on what’s the best measurement. Is it revenue? Is it margin? Is it signings? Is it churn? Is it renewal rate? Is it NPS? A whole set of different metrics that you can choose from with your executive team that help shape the conversation of your business because on one end of the spectrum, the kind of mission or operating principles speak to how you treat one another. The data is to help you kind of ground yourselves and how you’re doing as a team delivering on that promise.

Derek: Then the frequency with which that is shared. In my career, I’ve shared everything with salaries and every setting I’ve ever been in. So I’ve always had this kind of ultimate transparency theme which is I want everyone to be aware of what’s happening, why it’s happening, and then most importantly, what we’re doing to fix it and who’s responsible for fixing it because again, I don’t think this market accommodates people that want to kind of isolate and hide. I think you’ve got to kind of run in and use data as a way to level the playing field and invite everyone into the conversation, that the world, I think, perhaps our parents grew up in was the mushroom school of management, which was you keep them in the dark, you feed them shit and you can them. I think the world we live in now is transparent, it’s driven by trust and teamwork. That is how you are successful in this day and age.

Kirill: Got you.

Derek: Data’s the epicenter of that.

Kirill: I agree. Is that why you have this phrase where you say data will become the new operating system?

Derek: Yeah. Well, I think data is the new operating system is, it speaks to a number of elements. one is, it’s how you’re absolutely going to run your business, that is the underlying kind of back plane of your business. I think the other part to it is that you need to create an environment where every bit of data that’s created or that’s onboarded into your environment has to be aware of and cataloged and secured and governed and enabled for everyone that’s within the business. And you only get that by designing an operating system that facilitates it, right? So every application you build is built on that data operating system.

Derek: So let’s say the marketing group decides to run a campaign. They pull data through their marketing automation software. Well, that data should be also sitting within the data operating system so that the next time the sales team or the financial organization or the support organization understands there’s a client issue, they can look at the complete lineage of, they can do subsequent analytics of all the different ways in which that campaign touched that client, get a 360 review of your business. You don’t get there unless you have a common operating system with which all those applications run on.

Kirill: Got you. Would you agree with the statement that in the next 10 years every person in every organization is going to be expected to have some level of data literacy?

Derek Schoettle: Yes, a hundred percent. Well, I’ll say it this way. I think the ones that win will drive that. I think the ones that accept mediocrity, if not lose, are the ones that just don’t have that perspective and are not willing to drive that, kind of make that the norm. I think that’s the big risk.

Kirill: Got you. All right. We’re slowly coming to the end of the podcast. I had a very philosophical question for you. From all the experience that you’ve had in the multiple companies that you’ve worked in and headed and especially through IBM, where do you think the field of data science is going and what should our listeners prepare for to be ready for the future that’s coming?

Derek: It’s a great question. I mean, I would say it’s coming out of the shadows. I think the quote-unquote number of people that are going to be expected to be working with data in the context of science and application development is going to explode. I think the real opportunity for folks that are working in this field today as they look out into the future is how they can help shape how data and the various tools that are used and the platforms that are deployed on, how this market takes shape.

Derek: I think the really exciting part, especially now with self-service and Github and StackOverflow and various forums where people are now … I mean, this podcast. You can easily connect with and communicate with such a broad set of folks that I think helping influence where this market’s going and the rate at which it expands and invites a bigger seat at the table, back to our data culture question, is the opportunity, right? We have now the emergence of the chief data officer, right, whereas five years ago that really didn’t exist, and if it did exist, it was in the context of risking compliance, not in how do we build new business models, how do we monetize the data that sits within our business and our broader economy. I think there’s a chance for this to kind of become a centerpiece at any executive slate where they’re helping shape the businesses’ strategy, they’re helping dictate where and how and when the company’s spending money to reach new clients, what products are built. I think there’s a real chance for this discipline to take a far bigger role in every aspect of business.

Kirill: Very interesting, you mentioned executives. I actually had this other thought. Do you think that executives in this day and age of different levels, CEO, CFO, COO need to undergo some sort of data training-

Derek: Yes.

Kirill: … and have an understanding? If so, to what extent?

Derek: Yeah. I think everyone … There’s one dimension which is from business compliance and risk and awareness, just understanding just how data can be harnessed, how fast it’s growing, the range of types, meaning, unstructured, semi-structured and structured, that the range of technologies and how people can take advantage of it, that then speaks to value, right, in terms of better offers to your clients in real time, understanding what decisions to make about different investments, lots of different scenarios there. I think there’s that dimension. I think the other dimension is understanding just how quickly your business can change if you’re willing to embrace the technologies and the subsequent kind of investments you can make if you harness the data in the right ways. By that I mean, do you really understand the various open source tools? You don’t need to know how … You don’t need to go on Kaggle and win a competition, you don’t need to go into a Jupyter notebook and understand how to use Python, but I do think you need to understand that this is the new kind of landscape.

Derek: This is where this next set of these teams and these investments and these discussions are going to lead us, and if I’m in an executive capacity, you have to understand that. You can’t just assume that it’s going to be sitting in a kind of static enterprise data warehouse, perhaps with a few teramarts off to the side and some reports that are emailed out every Friday to let how you did in the previous month. I don’t think that’s the world we live in. I think the world we’re heading into is going from batch to real-time, going from, I’ll say, cloud to edge and from singular to team and from internet scale to, I’ll say, whatever 10x internet scale is.

Derek: I just think the rate at which we’re creating data … That’s why AI is going to play such an important role in this whole discussion, right? I mean, we are really focused on AI for business because we really do believe it’s what’s going to allow whole industries to transform themselves, whether that’s health care, whether it’s financial services, whether it’s retail, logistics and shipping, all of it is up for grabs. We just did this big conference in the beginning of the year called Big Bets where we got a hundred CEOs together and our CEO, Ginni Rometty, and a series of our senior leadership team to talk about, this collection of executive leaders, what do they see coming? And the number one thing was data and the impact on their businesses. So if we everyone agrees that it’s going to change the way that our business is collectively run, then yeah, you better have a level of competency to that then informs how you run your business.

Kirill: But the question then is how? I’m sitting here as, let’s say, I’m an executive listening to this podcast, and then I’m like, I really want to do that, but how do I get that data leaders?

Derek: Yeah. Great examples of what we’ve done at IBM is we partnered with Galvanize, which is a company here in the US that has a global footprint to create curriculums for data and science so that it can be for the developer, it can be for the business analyst, it can be the executive and that goes through everything from here’s the current, I’ll say, solutions, here’s the business models, here’s the different ways in which you can think of data and the technologies that can help you take advantage of it. I think there’s another dimension, which is your peer set. I spent a lot of my time talking to contemporaries of mine in other businesses that are earlier stage, smaller in scale, but having conversations with them about what applications they look at. Do you have an account with Kaggle? Do you have something like Data Science Experience? Do you have an account where you’re able to use and access data to make decisions, meaning, do you have something like Watson analytics? Do you have something like a whole selection of different tools? What are you using and how are you staying contemporary? You’ve constantly got to be learning. You just you always have to be learning.

Kirill: Fantastic. Thank you. Wonderful insight. I think executives listening to podcast are now satisfied and have a lot to look into. Well, Derek, it’s been a pleasure. Thank you for coming to the show. Quick question. Where can our listeners contact you, follow you, find you, follow your career or get more about these amazing things that you guys are doing at IBM?

Derek: Yeah, no, LinkedIn or Twitter at dschoettle, and usually those places find me pretty quickly. I think what I’d always encourage is if you go to IBM.com, we have put … I’ve been here four years, and we’re five years into and I think just completed our transformation back to growth. The digital presence for all of this technology is meant to be inviting. It’s self-service. You can remain anonymous. You can try out these new technologies on our platform, and let me know what you think. You can hit me up directly.

Kirill: Thank you very much. One final question for you today is what is your favorite book that you can recommend to our listeners to help them through their careers?

Derek: Yeah. I have a lot of books, but I think for this conversation in particular I would read Stanley McChrystal’s book called Team of Teams. It speaks a lot to how we think of our clients and where we see them going and certainly where we’re going and operating as it relates to data and how our teams are structured, how they’re motivated, how they’re incented. I think for anyone starting a business or running a business of I think any scale, but certainly north of several hundred employees would find it incredibly beneficial. I know that I have.

Derek: I would also say another book that I have recently read that I did so as a result of being at IBM, and this is a company that’s 106 years old, it’s had a number of different stages and its growth and its transformations around the mainframe and around the PC and now around the cloud, but there’s a great book about the founder called The Maverick and His Machine about Thomas Watson, Sr. It really tells the story of the founding of IBM and the culture that he created, and which is still very much alive today, that in my short time I didn’t fully appreciate it until I read that book, but now that I’ve read it, it is incredibly encouraging because what he talked about was we’re building a business that is founded on trust and on transparency, we’re building a business focused on clients and client outcomes. It just makes you understand IBM as a company. I think going into this next decade that’s going to be dictated by AI and data, I think those companies that can be trusted and that can be looked to as a company doing good in the world is incredibly important, I think, now more so than ever. And this book will lay it out for you why IBM is who it is.

Kirill: Fantastic. So there you go, Team of Teams and The Maverick and His Machine. Once again, thank you so much, Derek, for coming on the show. It’s been a great pleasure, and thank you for sharing all these insights. I’m sure so many people are going to find them valuable.

Derek: Well, I hope so, Kirill. Great chatting with you, and hope we can do this again sometime.

Kirill: So there you have it. That was Derek Schoettle from IBM, the chief business officer of the Watson and cloud platform. Hope you enjoyed today’s podcast. As you can see Derek spared some time and wasn’t holding back at sharing all of the insights, all of the latest updates and all of the latest progressions in the space of cloud platform these days. I really wonder what your favorite part was. For me, personally, it probably was the whole notion of data science being a team sport. It was an interesting idea. I’ve never heard it being put that way before, but if you think about it, you got to agree because we are moving into a world where more and more data is coming in all the time and you want more and more people to have access to it. You want that self-serve analytics component in your business to thrive, and you want everybody to be educated. That’s exactly what we’re talking about, and cloud enables that. Derek demonstrated that with some very good points and examples. Is your business or is your team treating data science as a team sport? Something to ponder on.

Kirill: If you enjoyed this episode and you know somebody who’s in charge of a business or who is looking to build a data-driven business, then make sure to forward it to them. On the flip side, if you are interested to learn more about the IBM cloud platform, about the Data Science Experience which they launched just over a year ago, then make sure to hit Derek up about it. You can find all of the links to the materials that we discussed at the show notes at www.www.superdatascience.com/137. There you’ll also find the URL for Derek’s LinkedIn and Twitter, and you can connect with him there, I highly encourage you to do that to stay up-to-date with everything that’s going on in the cloud. I was really excited to have you on board here today. Can’t wait to see you back here next time. Until then, happy analyzing.