 Live from Las Vegas, it's theCUBE, covering AWS Executive Summit, brought to you by Accenture. Welcome back and good morning. Welcome to theCUBE's live coverage of the Accenture Executive Summit here at AWS Reinvent. I'm your host, Rebecca Knight. We are joined by Prasad Sankaran. He is the Senior Managing Director, Global Lead, Intelligent Cloud Infrastructure. Welcome back on theCUBE. Thanks so much for coming on. And Larry Socker, Global Managing Director, Infrastructure, Services, Domains, and Strategy. Thank you so much. Good to be here, Rebecca. So we're talking today about hybrid cloud, the best of both worlds. As we say on theCUBE, it's a multi-cloud world. But I want to start with you, Prasad. What is driving clients to move to the cloud? What are you hearing from CIOs? What keeps them up at night? So I think, you know, we've got to a point with our clients where they're really trying to get the power of the cloud to the enterprises. So there are multiple things that they're trying to do. First, they're trying to really get the innovation that cloud provides in driving their digital transformation. The second is taking advantage of the cost savings that cloud can provide. So there are multiple aspects of how they use cloud. The first would be using SaaS-type applications. So for example, Salesforce.com or Workday and things like that. The second is using the power of providers like AWS really to drive what they can do from a cloud-native perspective in building new applications. And the third is just taking existing applications around legacy and either re-hosting them or refactoring them as in adjusting them to some extent and then hosting them on again, clouds like AWS. So that is a multi-year journey that all our clients are on. And you know, as Accenture, we're helping them identify what needs to go first, what needs to go next and help them in that journey. And what is the compelling force? Is it that they want to save costs? Is it that they want to be more innovative and they view this innovation in the cloud as the key to it? I think it's a combination of both. It's innovation is absolutely number one. Secondly, it is speed to market, the ability to get your product out there very quickly is second. And the third is at the end of the day, data centers are going to go away and we're going to all reside in the cloud in some form or fashion, whether it's public or private and which is why this whole topic of hybrid and multi-cloud is becoming so important today. So Larry, why hybrid cloud though? Why can't we just go all in on public, sorry? It's interesting. The real driver, I'm touched on it, the real driver to get to AWS and to the hyperscalers really is around the innovation cycles, the passes, the services that they can do and that drives innovation, the speed to get there is important. Gives you a way of quickly scaling, which if I prefer everyone to build on an application fast, it's a great way to get there. And there's obviously the consumption economics. How do I shift from CAF-X to OPEC? So that's driving the cloud native, the push into the world of public. At the same time, our clients do have a number of requirements that really make them look and rethink and figure out how to evolve their data centers. The first ones were regulatory. So you think about when you were the pharmaceutical, GXP compliance, HIPAA and the healthcare side of things, GDPR. For every company on the planet. Yes. So that regulatory was perceived as a barrier, particularly in some of the more difficult regulatory environments. And now, while the public providers are really evolving and starting to get better regulatory postures, at the same time, a lot of our clients were making investments and decided, hey, how do I really build my private cloud? Another big driver that people continue to look at private clouds, whether they're in their data centers or increasingly moving to colas, is just scale up architectures, like high terabyte HANA. If I want a 64 terabyte HANA deployment, while the AWS footprints get bigger and bigger, sometimes just the performance and tuning for those high scale data environments is a big deal. Oracle, Oracle Rack's a great example, where it's not only do you have something a highly tuned environment, but given some of the licensing arbitrage stuff, it makes it extremely difficult. And then a big part of it is data gravity. If I've got these big data sets, if you think about fraud analysis and Hadoop clusters for credit card processing, or where I've got high terabyte HANA database, I've got 20 or 30 applications that need to access that data, I can't really put it over a wide area network due to latency and the cost of moving data around. So what you ultimately end up with is applications clustered around lakes, big pockets of data. And I think that's where we're ending up. And that will be across a hybrid architecture. So that's what we call it. As you look at balancing your apps and your data across public, private solutions, that's why we view hybrid as the way that most of our clients are going, given the scale and the amount of applications and data they have. So that's what we refer to as the best of both worlds for hybrid. So I saw you nodding a lot with what he was saying, Prasad. How do you, as you said, it's a balancing act in terms of how you set your strategy. How do you recommend companies go about thinking in terms of how they allocate their cloud? Yeah, so I think we have to take a really, really take an application-centric approach. You know, I'll go back a couple of years when our clients were really looking to really use a public cloud, and then they've signed up with one or more public cloud providers, and then they move some applications. And then some of them have actually taken a step back. In fact, there's a very global investment bank that I've been working with who had taken that approach initially. And now what they've done in working with us is taking a very application-centric approach. So we study their entire suite of applications, understand from a regulatory perspective, from a compliance perspective, from a perspective of security. And this is a global bank, so there are different roles in Europe versus in the United States and so on. So based upon all that, we come up with an approach on what should reside in a private cloud as opposed to what should reside in a public cloud. And obviously there are multiple providers. There are reasons to go with more than one cloud from a public perspective, et cetera. So we advise them on that. Then once we're able to do that, then we chart the journey on what application gets moved and when. And certain applications are very important to them from a performance perspective and they need to scale up and so on. So in those cases, we treat them differently. In certain other application cases, we move them on to a pass. And in some other cases, we just move these applications into what's available from a SaaS perspective. So really it's a very much an application-centric approach on where the workload's at. And is it a living and breathing thing? I mean, we talk about cloud being this journey, it's not a destination. Does this change over time? I mean. It absolutely does. I mean, it's constantly evolving. You know, data patterns are coming out there and it was interesting. Like a couple of years ago, all we talked about was the app, how you're going to modernize the six or seven hours when you get reimagined in there. But all of a sudden, a lot of the conversations shifting to my data strategy. Where did this data reside? Particularly as the data sets get larger. And that's moving from what was very centralized in the public data centers into much more distributed architecture. So we see it evolving very quickly. And even with a single application, I got a great example with a global hotel. They had a reservation system and a strategic application running on an old IBM mainframe. You know, they were finding it was just taking too long to get innovative and agile to really support their new mobile applications, their web channels. So, you know, they looked at the, hey, if I was to reimagine, rewrite this, you know, go put it up into Amazon, you know, how long will it take? You know, how much would it cost? They found it was going to be two or three years and they didn't have the luxury to wait. So they basically wrapped that application with APIs. They exposed it, you know, as with microservices and then developed a cloud native front end with a database cache, new technologies that they can now drop every three or four days. They can now, you know, you get a new mobile application. So got agile delivery. They've still got the legacy stuff there. The data is still there. Now, when you go against their web and mobile application, you're actually browsing against that new cloud native, you know, the database cache when you look at rates and rooms, et cetera. It's only when you transact and reserve the room that it goes back to the mainframe. So you agile delivery, they get a lot of the benefits there and then they can, you know, just offload a lot of the processing on the mainframe. It's only really, you know, all the reads up front now and slowly deprecate that over time. Now, they, you know, they're now running that very interesting hybrid deployment architectures. They have a container approach, you know, this database cache is they have two clusters of those containers running on private clouds, you know, on VMware and then one running in AWS. So they're now can optimize, you know, across the public and or hybrid footprint, you know, how they do this. So over the time, they then look to evolve that application and start to introduce serverless. So starting to take advantage of Lambda. So there, even within a single application, you can see, you know, it's constantly evolving. They're starting to advantage of the next evolution. So the next one to move, you know, beyond containers into more serverless architecture. So, I mean, you both have just given me terrific examples, you're in, yours in financial services, yours as a hotel chain. How do you develop best practices? I mean, when you're working with clients, or really is it case by case? I mean, you talked about the decisions and the allocation of cloud, of course, there are regulatory constraints for each company, but are there industry-wide best practices? Yeah. Absolutely. I do want to. Sure. So I'll start with financial services as an example. So if you take the world of banking or insurance, there's obviously, you know, regulatory requirements, compliance requirements that they have to look at. And also if you look at, for example, in banking, a lot of it is very mainframe-based. You know, they have very old core banking systems, which is not possible to really move all of that onto public cloud necessarily. And maybe the business case isn't even there. So we have to take a digital decoupling type of approach and figure out, you know, how do we take some of the best bits and then put that on public cloud while still keeping some of the, you know, major amount of data still on the legacy side. So I think it's really an industry-specific approach, on top of which you have to layer on what the local requirements are. Yeah, absolutely. I mean, so I've spent a lot of work with Life Sciences and Pharmaceuticals and, you know, GXP compliance rules the day there. Yeah, you know, so how did they start to, they started off, you know, how do we get to private clouds that look and feel a lot like public and then start to move and test the waters, you know, as they start to migrate into the AWS's of the world. So it is very, you know, industry-specific. So we basically developed a library of, you know, this is how, you know, this is how you solve for banking, you know, the retail banking on a mainframe, how do I do similar pattern to that hotel front-ends? You know, insurance is actually very similar, too. A lot of the logic on mainframe, it doesn't make sense to rewrite, there's not enough value in that. If I can wrap it in microservices and move out. So that clouds, I mean, their roadmaps, now obviously every, given the scale and complexity of our clients, everyone's going to be slightly different, but their patterns are very industry-specific and it gets getting even more interesting as we start to get and move into IoT and Edge, because when you get into like the, you know, the connected oil and gas or connected mine, you know, all of a sudden it becomes extremely verticalized and industry. So the hybrid solves a lot of problems, but I imagine it also creates some other challenges, too. What are the challenges and how do you counteract them? Do you want to start, Larry? Yeah, sure. I mean, and it's not just hybrid, I think it's, we have much more complex architectures. So with all the power of digitally decoupling, you know, creating microservices, architectures, being able to pick, you know, best-of-breed services, as everything becomes much more dynamic and ephemeral. So we're moving for a world where a server and a virtual machine would be up for like 12 months or 15 months to containers that have lifespans of minutes or hours or even seconds, you know, now managed in a much more dynamic way with, you know, Kubernetes. So that hotel is 15,000 containers managed by Kubernetes optimized over that environment. So, and even more with serverless. So you've got a very complex environment to manage. And I think as our clients, you know, start to really evolve the application portfolios, you know, with SaaS, cloud-native, wrapping legacy, the ability to then seamlessly manage across that environment and optimize. And even the optimization, you know, in the old days, we used to optimize around one or two dimensions, was it cost or performance in SLAs. Now we've got to simultaneously take this very complex environment and optimize across, you know, performance, service levels, security, compliance, and cost. So it's not just about cost optimization and they offset each other. If I'm optimizing for, you know, performance and service levels, my cost probably goes up. So it becomes a very complex problem. So spend a lot of time looking first, you know, first starting with the operating model. You've got to operate differently. How do you really get DevSecOps involved so that operations and security baked in right when you're doing the analysis so you're not building something that can't be secure or operated? You know, and then really transforming the people, right? I mean, one of the hardest things, a place where our clients struggle the most, is how do they, you know, upskill their organization, change the culture of the behavior? Because, you know, great example is we look at how we operate cloud and infrastructure. You know, we want to turn our operators from, you know, eyes on glass, you know, looking at consoles into developers who are writing the next analytic algorithms to figure out predictive operations, who are doing the automation script to take, you know, to remove mistakes, you know, streamline, you know, drive agility and reduce cost. And ultimately, to tune the AI engines that are going to need it to do that complex optimization across very, you know, very complex architectures. Well, so you're talking about the people and the change management. I want to ask about innovation within Accenture and AWS. We heard Andy Jassy this morning in his fireside chat talk about how the company is now, obviously a ginormous company, but how it really still has this startup mentality and what he does to ensure that innovation is still a priority, relentless focus on the customer, decentralized nature of the organization, really focusing on what you're good at, knowing what's in your wheelhouse. How do you think about innovation and then how do you help clients make sure that they are bringing people along? And as you said, giving them the sort of developer mindset. Do you want to? Yeah, sure. So that's absolutely important for us. Innovation is, you know, very much part of our culture. When I look at my own group, which is intelligent cloud and infrastructure, what we do is we have 30,000 people who are working every day at our clients. And in many ways they understand the client's landscape even better than anybody else because they're working there every day. So some of the things that we do with our clients are innovation forums where our people bring forward ideas to suggest to clients. So this is something that we do on a regular basis. Take it to our clients, CIOs, and invariably several of these ideas actually get put into their whole plan for the following year. So we co-innovate with our clients. We've created several digital studios and liquid studios in multiple locations. We've got several of them across the globe. So we bring our clients in for design thinking workshops, suggest ideas, and, you know, co-innovate with them on things that we can take forward. And just to further that, I mean, we've always driven an innovation agenda. We're always trying to figure out what's next, where do we invest at, you know, what's the next bet? I mean, even going back to our cloud first days, I mean, so we've always been pushing the envelope. I think what Prasad hit on is really a step change for us in how we engage. I think in the past, we'd come in as the consultants, present, et cetera. And I think, you know, our clients are bright people. So actually engaging with them in more design thinking. We acquired a bunch of digital studios, Fiora, and stuff. And I think they've infused this culture of, you know, co-sourcing, collaboration, and ideating together. And the dynamic just changes. And it's fantastic when the client then is in the room, they own it with you. They co-creating. And I think, you know, so we've been spending a lot of time on how do we transform, how we interact and engage with our clients in a much more collaborative way. And it's really changed, you know, changes your relationship with them. And the technology enables that too. Absolutely. Yeah. Well, Larry and Prasad, thank you so much for coming on theCUBE. A really great conversation. Thank you. Thank you, Rebecca. I'm Rebecca Knight. Stay tuned for more of theCUBE's live coverage of the Accenture Executive Summit coming up in just a little bit.