 Okay, welcome to theCUBE. We're live in San Francisco. This is theCUBE at the GE Industrial Cloud Event, hashtag Industrial Cloud. Tweet us if you have any questions. I'm John Furrier. This is theCUBE, our flagship program. We go out to the events, extract the ceiling from the noise, and I'm joined by my co-hosts. I'm Dave Vellante, wikibon.org. Paul Daugherty is here, he's the CTO of Accenture. Paul, welcome to theCUBE. Thank you. We think Accenture, thanks for making some time. So we think Accenture, we think deep industry expertise, deep knowledge, hands-on, like-hard problems. So this is a big, chewy, deep expertise problem. Talk about, first of all, what you do at Accenture, and then we'll get into the whole relationship with GE and the opportunity that you guys are going after. Yeah, sure, yeah, good question. So I've got the role of chief technology officer within Accenture. So what I'm responsible for is our technology labs looking ahead to the next generation of technology and how we use our technology labs and our research and development to bring that back into our business. I'm also responsible for the emerging businesses that we start up in areas like cloud, like mobility, like big data, like social, so those areas report to me and as we look to integrate those into our core business. So are you kind of like a chef looking at the best ingredients out there? You guys don't, well, I guess you do make some technologies, but you're not a software player. I mean, you write software, but you're not competing with the IBMs and the HPs of the world. I mean, you do compete on the services side, but you're not tinkering and developing tech per se for mass distribution. So are you more like a chef looking at the best technologies out there or are you guys increasingly developing your own tech? It's a little bit of both. So our position in the market would characterize as independent with a point of view. So we work with the technologies that are going to deliver the best value for our clients and that involves a number of vendors, a number of tech companies, and it varies by industry, varies by company based on what they want to use. So we have to be ready to work with whatever technology is right for the company, but then within certain industries, within certain segments, within certain geographies, we can see the patterns and see what's going to add most value to the customer. So we'll have a point of view that if you're trying to solve this problem, then this kind of technology is going to help you do it most effectively and we'll then invest in the architectures, accelerators, toolkits, training for our people, et cetera, to deliver more value to our clients around those particular technology platforms. So, CTO, you do a lot of strategy. You probably do a lot of due diligence. And as you say, you put together those architectures so that you can actually develop solutions because that's what your clients care about. So take us back to when you first started investigating this so-called industrial internet. What were some of the texts that you were looking at? What's your vision? And then what's the platform that ultimately you guys are building out? Yeah, the good question. So what led us to GE and industrial internet and the discussions we're having was a broader phenomenon we're seeing happen around digital business that I talked a little bit about during the session. And we are really seeing this big momentum shift toward digital business across industries. And it started in retail and industries that are more consumer oriented with companies shifting their model. They're shifting to digital marketing, digital commerce, et cetera. But when you look at how to apply that and when our clients look at how to squeeze the digital value out of other parts of their business or capitalized in digital opportunities, they're impeded by the fact that they're manufacturing the physical part of the business that they run is often outside the scope of their IT systems. It's not connected, it's not online, et cetera. So we looked at how to deliver value and how to achieve that digital vision for companies that operate physical types of industries or B2B types of industries. That led us to this notion of industrial technology and the convergence of industrial and operations technology with information technology. So it's IT plus OT come together with information technology. So we've been looking at that for a while. We acquired a company a number of years ago that was in the industrial technology area that was in refinery automation. And we explored that and built a business around that. We built a connected vehicle platform that we've used with a number of the automotive companies to help them build their connected car and infotainment and navigation services and those types of things in the car. So we've got a number of steps into the industrial world through those, you know, by working in some of those early adopting industries. So the real-time nature of this opportunity is somewhat unique. I mean, everybody talks about real-time and in the mainstream world, but this has really got to be real-time, machines making decisions to figure out how machines should operate, how hard you should push a turbine and how to optimize that over time. So, but at the same time, you've got all this historical data we heard about, you know, the sort of time machine that we heard about today. So how do you marry those two? What kind of requirements, unique requirements, does that real-time entity put on architectures and how you guys behave? Well, it's a stretch, I think, for a lot of the, certainly a stretch for the architectures that our clients operate today, which is why I think there's a big opportunity to look at how the industrial internet evolves and what kinds of new technologies, new platforms that approaches our clients will need. The reality is there's not one answer. There's not one data schema. There's not one data architecture that's going to solve medical equipment problems in the health industry and refineries, automation and drilling problems and transportation, et cetera. So the challenge will be about having a flexible platform and understanding the data historian types of technologies for time series data, real-time machine-to-machine communications for data, smaller data that you need to get very quickly and Hadoop and technologies like that that can process massive amounts of data in memory technology, all the different tools that are available to assemble the right architecture in a given industry for a given company to solve the problem. So when we think about the platform and think about the industrial internet, it's about coming up with the right architectures and the smart engineering to solve specific problems for clients and it needs to be, you know, one size fits all platform, but a platform that allows the blending and engineering of different techniques. I mean, that's the challenge that we see too is, I mean, it's really, it's a conflict. I mean, you have, you want agile platform, but you also want to be able to vertically integrate and use those expertise and verticals, you know, and, you know, telematics to aerospace to wherever you still got to have the tooling at the top of the stack or the vertical. So how do you look at the developer environment? Because the developers are going to be building the apps and obviously, Marissa's got an interest in there with Pivotal and as well as EMC and VMware. You got AWS has been very successful with the developer community on the consumer side. So how do you look at that app developer, the operations productivity or the industrial productivity and then the infrastructure? Because you can flatten the infrastructure out and use virtualization, then as you start moving up into operations, how do you create that horizontal foundation technically? And what do the stacks look like? And this is what people, you know, the geeks want to know is, okay, if I go vertically integrate, am I going to be foreclosing an interoperable opportunity in another way with the data? So is that the data fabric that's open? So these are the kinds of things I want to get your thoughts on that vision. I know it's not baked out yet in the industry, but how do you look at that challenge? Yeah, I think one of the things we share amongst the partners that we're up on stage today, if you look across Accenture, Pivotal, Amazon, GE, is a view of how the stack needs to work in this environment. So I think the stack question is a very important one. So when you look at the stack at the cloud level, Amazon provides great capability to public cloud, everything that the Verner Vogels talked about. But clients are going to choose different solutions. Some of our clients will want hybrid clouds, some will want private clouds and other types of technology. So there's a cloud layer, but then there needs to be a layer above the cloud to say how do we move things across different platforms so I can do what makes sense in Amazon and Amazon, I can do what makes sense in my private cloud and my private cloud. So there's an abstraction layer above the cloud, and that's where technology that Palmer, it's in Pivotal has gets interesting in terms of managing that environment, as well as something we're bringing to the table called Accenture Cloud Platform, which is our management orchestration layer above the cloud to help clients manage and orchestrate. And the process of the whole environment. Is that where your versatility comes from? Is that where you guys are looking at saying, hey, I'll create an orchestration layer so that if someone in oil and gas wants to do something innovative, they can program away. And if someone over here in telematics wants to do something. You know what I'm saying? So it's like. It needs to be at that cloud level, you need some of that interoperability, and even for one client at one point in time, they're going to have multiple, for one application, they're likely going to have some things on a public cloud, some in a private cloud, some behind a traditional firewall. So it's managing a hybrid environment for one process in real time. That's the kind of problem we're focused on, the architecture for that. Then I think when you get to the other layers to your question on skills and tools and such, the data integration layer is a very important layer. And one thing we're looking at with the platform and the way we assemble it is how we bring in machine to machine communication and the right integration platforms to ingest and standardize the data in the right way. So one of the things we're working with as part, or we're bringing in and working with on this partnership is our Accenture Mobility Managed Services Platform, which manages machine to machine and data ingestion across a variety of platforms. So that gets into a standard data environment and there's a certain standardization across different technologies that you need there to help developers. And then the other layer where a number of technologies come together is a common analytics platform. So you can use the multiple data sources that have come together and come up with analytical models that span across the different technology, the different source data that you've got. And allow data scientists to learn from others models and advance them and deliver better outcomes for companies based on the evolution of better analytic models. And so that's the highest level of tooling and standardization that sees that ability to create those models. And that's a vision that Pivotal's embarking on with their data fabric and the way that they're creating their vision, which is one of the things that's interesting about the partnership and how it comes together. Is it reasonable to expect that those standards will emerge across industries or will we have like PACs? Industry-specific standards, data standards that emerge? I think there's some standards that will evolve based on certain domains and certain industries and there's some out there already. I think there's going to be a lot of problems that need to be solved on an industry basis. So the way you're going to get patient care data for a collaborative care and a healthcare environment is going to be very different than the way that you're going to get drilling data or mining data from those other industries. So I think there need to be multiple models in terms of how to pull together, which is why that data platform is very important because it needs to be very modular and pluggable so you can plug in new data sources from as you have different needs and different industries. When you say data platform, you're talking about like a Pivotal data platform or you're talking about something that you actually develop for clients. I think a data architecture. I think it's more of it. It's a data architecture. I think there's tools and architectures like Pivotal that are very important in that. But I think there's other capabilities that you need then as you look at how do you, what's the meta model look like in terms of the different source data you need? How do you integrate that in the right way and pull together? So I think there's some other master data and other metadata challenge that we saw. So the source data, the type of data, the metadata that you need to capture, the outcome that you're trying to achieve, that's really where you guys shine. Yeah, and the other area that I think is important, yeah, that's right. The other area that I'd add to that is the API layer of how you connect things together, which is an area where there's a lot of great innovation happening around the industry now in terms of different API and APIs as service models. And I think that API layer is going to be very important in this, that you can have a standardized common way of plugging things together at multiple levels in the stack rather than every time you have to solve a problem, doing the one-off point-to-point integration. And when you say that, Paul, you're talking about the API that a, for example, a Pivotal or some other open stack publishes, correct? Or are you talking about ones that you actually promulgate? I think there's multiple layers of APIs. There's the APIs going down into the stack and then there's APIs across application platforms. How do you connect your machine-to-machine data coming from a smart machine back to your SAP data, if that's what you need to do as one example? So the APIs across that layer of application technology will be very important. And that latter example is one that Accenture would actually develop or not necessarily? Yeah, we would develop the kind of end-to-end view of that working with some of the other partners and working with the industry that's being created around these API platforms. The API could connect the APIs. It's not complicated, is it? Yeah, well, you're connecting a lot of pieces together. So I think there's no equivalent. I think if you look back at prior generations of technology in the enterprise, with a back office you had a Waltz-Walt ERP, which simplified the environment for a lot of companies and they moved toward that. There's not that equivalent really out there in this area because it's more complex. There's more components you have, physical devices. It's not just a server and a data center. You guys love complexity. Well, we love making things simple for our clients. What our business is about is how do we make, create an outcome for a client that's simple for them to achieve? I mean, you guys have a track record of doing that. And that's one of the questions we were asking at earlier. Dave and I were talking on our intro, was managing complexity. You're talking about event processing, complex events. You're talking about network management. And these are things that were kicked around during the early wireless days. And if you go further back in history, in client server, network management, LAN, interoperability, internet working, TCPIP and moving packets around. So in a way that world was kind of stream-based event processing. So given that, what do you point to here, this industrial era that we're in, that they're talking here in San Francisco, the industrial cloud is really a new era. It's really a new thing that's happening. But what can you point to from the previous generations of IT that you've been involved in or technology saying, it's a little bit of network management mixed in with wireless packet management or orchestration software that's a middleware. I mean, you've got a kind of a cobble-together kind of paradigm here, right? It's a little bit of software engineering, you've got infrastructure. What does it look like? And what can you point to saying, we've seen this before? Yeah, I'd point to, in the seen it before category, we've thought about that. And I'd point to client server technology back in about 1991 when we're having lots of debates. Is it going to be LU62 or TCPIP? And is it Ethernet or Token Ring? Is it going to be CyBase or Informix? Is it going to be Fat Client or Thin Client? And that feels like where we are with the technology right now. And eventually, some choices became clear in terms of how you architected. Some different models became clear. And then technology standardized around that. You had packages that standardized and solutions that standardized throughout the stack. And I think what GE is trying to drive here with the ecosystem is accelerate that from the point of, you know, 1991 and client server to fast forward a little bit so we have the ability to do more standardized solutions. So I think it is, as you said, it's a lot of complexity right now. It's probably two, it's a lot more complex, I think that it'll be several years from now, some of the standards sort out. We get some of the base platforms in place. And that's why I think it's an interesting area. I think it's got a lot of expansion. It's uncertain too. And it's uncertain. Dave always talks about this and we always harp on it in that the practitioners make more money than the actual people supplying the technology. And we've seen that movie before in client server that for every dollar of SAP or Oracle, you'd see an X multiple distribution or just delivery. ERP, right? If you could have figured out who was going to apply ERP you could have made a lot of money. So with that context, we were building the foundation for the industrial internet and industrial things in the cloud. You guys are on the front lines. I mean, you're building your own software. You're becoming a software provider because by default that's the delivery requirement to simplify. So what's that going to look like to delivery? Who's the user? Who's deploying? What's the economics look like? Can you share any insight in there? Well, I think what we want to do for our clients again is we want to simplify the business outcome for them. How do we get them to their moment of business value quicker? And as you think about what cloud is, I think training the market to do is by based on outcomes and to buy more as a service. I think the way this looks at the end of the day is more clients buying the outcome. I think Bill, Bill Root talked about that in terms of how he sees the business model evolving. Clients will want more effective, more effective airline operations, intelligent airline operations. They'll buy outcomes around that. That's what we see it going as well. We talk a lot about everything as a service and clients increasingly buying an outcome based capability as a service. For us, we need to package the technology efficiently and package the solution efficiently so we can deliver it to our client on that basis. So I think that's the world we're moving into, I think. I think it's a good world from our perspective and from the customer's perspective because they get the outcomes they want. And from our perspective, we can solve a problem for them over a longer time horizon. It's not just about implementing technology. It's about helping them achieve the business model. And the big data piece is critical because you can actually measure the outcome and saying, did I achieve it? Yeah, right. It's like performance based solutions. It's transparency and accountability. You can see what happens. Well, thanks for coming on theCUBE. We really appreciate it. Paul Daugherty, the CTO of Accenture. Thanks for coming on. Great presentation. You guys are probably one of the most effective panels I've seen. I mean, I love the mojo of the industrial internet and the industrial cloud. It's got that futuristic, there's some tech involved that's evolving and dynamic. But also, you guys are talking about business operations and business value. And that's a conversation we want to hear more of. So thank you for coming on theCUBE. We really appreciate it. We'll be right back with our next guest after this short break. This is theCUBE SiliconANGLE's flagship program. Throughout the events, extract the season. I'm John Furrier of SiliconANGLE with Dave Vellante, Wikibon.org. We'll be right back.