 Live from Las Vegas, it's theCUBE, covering IBM Think 2018, brought to you by IBM. We're back at IBM Think 2018. This is theCUBE, the leader in live tech coverage. My name is Dave Vellante. I'm here with my co-host, Peter Burris. Moe Abdullah is here as the vice president of Cloud Garage and Solution Architecture, Hybrid Cloud for IBM, and Tim Davis is here, Data Analytics and Cloud Architecture Group and Services Center of Excellence, IBM. Gentlemen, welcome to theCUBE. Thank you. Moe, garage, Cloud Garage, I'm picturing drills and wrenches and what's the story? What's garage? I wish it was that top of a garage. My bill would go down, for sure. No, the garage is playing on the theme of the startup. The idea of how do you bring new ideas and innovate on them, but for the enterprise. So what two people can do with pizza and innovate, how do we bring that to a larger concept? That's what the garage is really about. All right, Tim, talk about your role. Yeah, I lead the Data Analytics field team and so we're really focused on helping companies do digital transformation and really drive digital and analytics and data into their businesses to get better business value, accelerate time to value. Awesome, so we're going to get into it. You guys both have written books. We're going to get into the field guide. We're going to get into the cloud adoption playbook. But Peter, I want you to jump in here because I know you got to run. So get your questions in and then I'll take over. Sure, so I think obvious question number one is one of the biggest challenges that we've had in analytics over the past couple of years is we had to get really good at the infrastructure and really good at the software and really good at this and really good at that. And we're a lot of pile of failures because if you succeeded at one, you might not have succeeded at the other. The garage sounds like it's time to value based. Is that the right way to think about this? And what are you guys together doing to drive time to value, facilitate adoption and get to the changes that the outcomes of the business really wants? So Tim, do you want to start? Yeah, I can start, because Moe leads the overall garage and within the garage, we have something we call the data-first methodology where we're really driving a direct engagement with clients where we help them develop a data strategy because most clients when they do digital transformation or really go after data, they're taking kind of a legacy approach. They're building these big monolithic data warehouses. They're doing big master data management programs. And what we're really trying to do is change the paradigm. And so we connect with the data-first methodology through the garage to get to a data strategy that's connected to the business outcome because it's what data and analytics do you need to successfully achieve what you're trying to do as a business? A lot of this is digital transformation, which means you're not only changing what you're doing from a data warehouse to a data lake, but you're also accelerating the data because now we have to get into the time domain of a customer or your customer where they may be consuming things digitally. And so they're at a website, they're moving into a bank branch, they go into a social media site, maybe they're being contacted by a FinTech, you've got to retain and maintain a digital relationship and that's the key. And the garage itself is really playing on the same core value of it's not the big beating the small anymore, it's the fast beating the slow. And so when you think about the fast beating the slow, how do you achieve fast? You really do that by three ways. So the garage says, first way to achieve fast is break down the problem into smaller chunks, also known as MVPs or minimum viable products. So you take a very complex problem that people are talking and over talking and over engineering and you really bring it down to something that has a client value, user centered. So bring the discipline from the business side, the operations side, the developers and we mush them together to center that. That's one way to do fast. The second way. But by the way, I did work with a client, they started calling a minimum viable outcomes. Yes, minimum viable outcomes mean product and there's a lot of these types of minimum viable to achieve and we're talking about four weeks, six weeks and so on and so forth. The story of American Airlines was taking all of their kiosk systems for example and really changing them both in terms of the type of services they can deliver so now you can recheck your flights et cetera within six week periods. And that's fast and doing it in one terminal and then moving to others. The second way you do fast is by understanding that the change is not just technology. The change is culture, process and so on. So when you come to the garage, it's not like the mechanic style garage where you are sitting in the waiting room and the mechanic is fixing your car. Not at all. You really have some sort of mechanical skills and you're in there with me. That's called pair programming. That's called test driven. These types of techniques and methodologies are proven in the industry. So Tim will sit right next to me and we'll code together. By the time Tim goes back to his company, he's now an expert on how to do it. So fast is achieving the cultural transformation as well as this minimum viable aspect. Hands on and you guys are actually learning from each other in that experience. Absolutely. And sharing, yeah. So I would also say I would think that there's one more thing but for both of you guys and that is increasingly as business acknowledges that data is an asset. Unlike traditional systems approaches where we built a siloed application, this server, that database manager, this data model, that application and then we do some integration at some point in time. When you start with this garage approach, data-centric approach, figure out how that works, now you have an asset that can be reused in a lot of new and interesting ways. Does that also factor into this from a speed standpoint? Yeah, it does. And this is a key part. We have something called data science experience now. And we're really driving pilots through the garage, through the data first method to get a rapid engagement. And the goal is to do sprints. To do 12 to 20 week kind of sprints where we actually produce a business outcome that you show to the business and then you put it into production. And we're actually developing algorithms and other things as we go that are part of the analytic result. And that's kind of the key. And behind that, the analytic result is really the icing on the cake and other business value where you connect. But there's a whole foundation underneath that of data and that's why we do a data topology. And the data topology has kind of replaced the data lake, replaces all that modeling because now we can have a data topology that spans on-premise, private cloud and public cloud. And we can drive an integrated strategy with a governance program over that to actually support the data analytics that you're trying to drive. And that's how we get at that. But that topology's got to tie back to the attributes of the data, right? Not the infrastructure that's associated with the data. It does. The idea of the topology is you may have an existing warehouse. That becomes a zone in the topology. So we aren't really ripping or replacing, we're augmenting. So we may augment a non-premise warehouse that may sit in a relational database technology with a Hadoop environment that we can spin up in the cloud very rapidly and then data science applications. And so we can have a discovery zone as well as the traditional structured reporting. And the level of data quality can be mixed. You may do the analytic discovery against raw data versus where you have highly processed data where we have extreme data quality for regulatory point. Do a God box where everything goes through some pipe into the box. And you put it on later. Well, Hadoop came out, right? People thought they were going to dump all their data into something beautiful was going to happen, right? And what happened is everybody created a lot of data swamps. Something really ugly happened. Right, right. It's just a pile of data. Or they ended up with a cheaper data warehouse. But it's not. The data warehouse was structured. It has quality. All the data modally, but all that stuff took massive amounts of time. When you just dump it into a Hadoop environment, you have no structure. You have to discover the structures. So we're really doing all the things we used to do with data warehousing. Only we're doing it in an incremental agile, faster method where you can also give access to the data all the way through it. It's not like we will serve no wine before it's time. Now you can eat the grapes. You can drink the wine because it's fermenting. No schema on right. Just throw it in. There is an image that Tim shows that the idea of a data lake is this organized library with books. But reality is a library with all the books dumped in the middle. Go find the books that you want. And they'll do just more. Exactly. If you want to pick on the idea that you had earlier, when you look at that type of a solution, the squad structure is changing, right? To solve that particular problem, you no longer just have your data people on one side. You have a data person. You have the business person that's trying to distill it. You have the developer. You have the operator. So the concept of DevOps to try and synchronize between these two players is now really evolved. And this is the first time you're hearing it right at the cube. It's the biz data DevOps. That's the new way we actually start to explain that. Very simple. It starts with business requirements. So the business reflects the user and the consumer. And they come with not just generic, you know, generic stuff. They come with very specific requirements. That then automatically and immediately says, what are the most valuable data sources I need, either from my enterprise or externally? Because the minute I understand those requirements and the persistence of those requirements, I'm now shaping the way the solution has to be implemented. Data first, not data as an afterthought. That's why we call it the data first method. The developers then, when they're building the cloud infrastructure, they're really understanding the type of resilience, the type of compliance, the type of meshing that you need to do, and they're doing it from the outset. And because of the fact that they're dealing with data, the operation people automatically understand that they have to deal with the right to recovery and so on and so forth. So now we're having this. You're not throwing it over the wall. Exactly. That's where the DevOps comes in. And you're also understanding the velocity of data, right through the enterprise as well as the gaps that you have as an enterprise because when you go into a digital world, you have to accumulate a lot more data and then you have to be able to match that. You have to be able to identity resolution to get to a customer and understand all the dimensions of it. Well, the digital world data is the core. So, and it's interesting what you were saying, Moe, about essentially the line of business identifying the data sources because they're the ones who know how data affects monetization. Inderpal Bhandari, when he took over IBM Chief Data Officer said, you must form partnerships with the line of business in order to understand how data contributes to the monetization. And your DevOps metaphor is very important because everybody's sort of on the same page is the idea, right? And there's a transformation here because we're working very closely with Inderpal State and the emergence of a Chief Data Officer in many enterprises. And we actually, we kind of had a program that we still have going from last year which is kind of the Chief Data Officer success program, right, where you can help get at this because the classic IT structure has kind of started to fail because it's not data oriented, it's technology oriented. So by getting to a data oriented organization and having an elevated Chief Data Officer, you can get aligned with the line of business, really get your hands on the data and we prescribe the data topology which actually the back cover of that book shows an example of one because that's the new center of the universe. The technologies can change, the data can live on-premise or in the cloud but the topology should only change when your business changes. This is hugely important. So I want to pick up on something Ginny Rametti was talking about yesterday was incumbent disruptors. And when I heard that, I'm like, come on, no way, instant skeptic, right? And so then I started to think about it and you guys, what you're describing is how you take somebody, a company who's been organized around human expertise and other physical assets for years, decades, maybe hundreds of years and transform them into a data oriented company where data is the core asset and human expertise is surrounding that data and learns how to look. It's not in, most data is in silos. You're busting down those silos and giving a prescription to do that. I think what Tim actually said is very, you heard us use the word prescriptive. You heard us use the word methodology, data first method or the garage method. And what we're really starting to see is these patterns from enterprises. You know what works for a startup does not necessarily translate easily for an enterprise. You have to make it work in the context of the existing baggage, the existing processes, the existing culture. Customer expectations, that scale, all of those have dimensions. So this particular notion of a prescription is we're taking the experiences from Hertz, Marriott, American Airlines, RBC, you know all of these clients that really have made that leap and got the value and essentially started to put it in a simple framework, seven elements to those frameworks. And that's in the adoption, yeah. So we got two documents here, the cloud adoption playbook, which Mo, you authored a little co-authored. With Tim's help. Tim as well. And then this field guide, the IBM data and analytic strategy field guide, that Tim, you also contributed to this. Which augments the book. So I'll give you the description of when I love the hybrid cloud data topology of the Mac. That's an example of a topology of the Mac. So that's kind of cool. But go ahead, let's talk about this. So if you look at the cover of that book and piece of art very well drawn, that's right. You will see that there are seven elements. You start to see architecture. You start to see culture and organization. You start to see methodology. You start to see all of these different components. Governance, management, security. That's right, that really are important in any type of transformation. And then when you look at the data piece, that's a way of taking that data and applying all of these dimensions. So when a client comes forward and says, look, I'm having a data challenge in the sense of how do I transform access? How do I share data? How do I monetize? We start to take them through all of these dimensions. And what we've been able to do is to go back to our starting comment. Accelerate the transformation. Accelerate. And the real engagement that we're getting pulled into now in many cases and getting kind of pulled right up the executive chains at these companies is data strategy. Because this is kind of the core. You've got to, so many companies have a business strategy, very good business strategies. But then you ask for their data strategy. They show you some kind of block diagram architecture. They show you a bunch of servers in the data center. That's not a strategy, right? The data strategy really gets at the sources and consumption, velocity of data, and gaps in the data that you need to achieve your business outcome. And so by developing a data strategy, this opens up the patterns and the things that we talked to. So without we look at data security, we look at data management, we look at governance, we look at all the aspects of it to actually lay this out. In another thought here, the other transformation is in data warehousing, we've been doing this for the past, some of us longer than others, 20 or 30 years, right? And our whole thing then was we're going to align the silos by dumping all the data into this big data warehouse. That is really not the path to go because these things became like giant dinosaurs, you know, big monolithic difficulty change. You know, the data lay concept is you leave the data where it is and you establish a governance and management process over top of it. And then you augment it with things like cloud, like Kadoop, like other things where we can rapidly spin up and we're taking advantage of things like object stores and advanced infrastructures. And this is really where Mo and I connect with our IBM Cloud Private platforms, with our data capabilities, because we can now put together managed solutions for some of these major enterprises and even show them the roadmap. And that's really that roadmap. It's critical in that transformation. Last word. Yeah, so to me, I think the exciting thing about this year versus when we spoke last year is the maturity curve. You know, you asked me this last year, I said Mo, where are we on the maturity curve of adoption? And I think the fact that we're talking today about data strategies and so on is a reflection of how people have matured. Earlier on, they really started to think about experiment with ideas. We're now starting to see them access detailed, deep information about approaches and methodologies to do it. And the key word for us this year was not about experimentation or trial, it's about acceleration. Because they've proven it in that garage fashion in small places. Now I want to do it at the American airline scale. I want to do it at the global scale. And I want to, so acceleration is the key theme of what we're trying to do. What a change from 15, 20 years ago when the data warehouse was the single version of the truth. It was like a snake swallowing a basketball. And you had a handful of people who actually knew how to get in there. You had this huge asynchronous process to get insights out. Now you guys are very important in the year. It's the marketization of data. Everyone should be. So guys, really exciting. I love the enthusiasm. Congratulations. A lot of more work to do. A lot more companies to effect. So we'll be watching. Thank you for coming. Thank you very much. And make sure you read our book. Yeah, definitely. Read these books. Cloud is Option Playbook. And IBM Data and Analytics Strategy Field Guide. Where can you get these? I presume on your website. You can get them on Amazon. Great. Okay, good. Thanks guys. Appreciate it. Keep it right there, buddy. This is theCUBE. We're live from IBM Think 2018. And we'll be right back.