 Live from San Diego, California, it's theCUBE. Covering Cisco Live US 2019. Brought to you by Cisco and its ecosystem partners. Welcome back to San Diego, everybody. You're watching Cisco Live 2019. This is theCUBE, the leader in live tech coverage. This is day three of our wall-to-wall coverage. We go out to the events, we extract the signal from the noise. My name is Dave Vellante, Stu Miniman is here. Our third host, Lisa Martin is also in the house. Dave Malick is here. He's a fellow in chief architect at Cisco. David, good to see you. I'm glad to be here. Thanks for coming on. First of all, congratulations on being a fellow. And what does that mean at Cisco Fellow? What do you got to go through to achieve that status? It's a pretty arduous task. It's one of the most highest technical designations in Cisco, but we work across multiple architectures and technologies, as well as our partners as well, and to drive corporate-wide strategy. So you've been talking to customers here. You've been presenting. I think you said you gave three presentations here. Multi-Cloud, Blockchain, and some stuff on Machine Intelligence, ML. Yes. Let's hit those. Kind of summarize the overall themes, and then we'll maybe get into each, and then we've got a zillion questions for you. Sure, excellent. So Multi-Cloud, I think one of the customers we're clearly hearing from them is around how do we get a universal policy model and connectivity model, and how do you orchestrate workloads seamlessly? And those are some of the challenges that we're trying to address at this conference. On Blockchain, a lot of boys out there were not talking about Bitcoin or cryptocurrency. It's really about leveraging Blockchain from a networking perspective on identity and encryption and providing a uniform measure that everything is pervasive across infrastructure. And then ML, I think it's the heart of every conversation, how do we take pervasive analytics and bring it into the network so we can drive insights and actionable insights that we can drive into automation. So when you talk about, let's start with the third one. When you talk about ML, was your talk on machine learning, did it spill into artificial intelligence? What's the difference to you from a technology perspective? So machine learning is really getting a lot of the data and looking at repetitive patterns in a very common fashion and doing a massive correlation across multiple domains. So you may have some things happening in the branch, the data set are a WAN and cloud, but the whole idea is how do you put them together to drive insight? And through artificial intelligence and algorithms, we can try to take those insights and automate them and push them back into the infrastructure or to the application layer. So now you're driving intelligence for not just consumers or devices, but also humans as well to drive insight. All right, so Dave, wonder if you help connect with us what you were talking about there and we'll get to the multi-cloud piece because I was at an Amazon show last week from Amazon talking about how when they look at all of the technologies that they use to get packages, their fulfillment centers, everything that they do as a business, ML and AI, they said is underneath that. AWS is what's driving that technology from that standpoint. Now, multi-cloud AWS is a partner of yours. Can you give us how you work in multi-cloud and does ML and AI, is that a Cisco specific? Are you working with some of the standards out there to connect all those pieces? Help us look at some of the big picture of those items. So we believe we're agnostic whether you connect to Amazon, Azure, Google, et cetera. We believe in a uniform policy model and a connectivity model, which is very, very arduous today. So you shouldn't have to have a specific policy model, connectivity model, security model, for that matter, for each provider. So we're normalizing that plane completely, which is awesome. Then at a workload level, regardless of where your workload is spun up or spun down, it should have the same security posture and visibility. Now, we have certain customers that are running a single application across multiple clouds. So your data is going to be obviously on-prem. You may be running analytics and TensorFlow compute an EC2 and connecting to all 365. That's one app. And where we're seeing the models go is you're leveraging technology such as Istio for service mesh, how we tie a lot of these microservices together, and then we have to layer workload orchestration on top. So regardless of where your workload sits. And one key point that we keep hearing from our customers is around governance. How do you provide cloud-based governance regardless of where the workload is? And that's something we're doing in a very large fashion with customers that have a multi-caller strategy. So Istio, I think there's still some confusion around multi-cloud generally, maybe Cisco strategy. I wonder if we could maybe clear it up a little bit. I mean, Steve, it's that big elephant in the room. And I always feel like everybody describes multi-cloud from a different angle. So let's dig into this a little bit. And let's hear from Cisco's perspective. So you got, to my account, five companies really going after the space. You got Cisco, VMware, IBM Red Hat, and Microsoft and Google, Google with Anthos. Probably all those guys are partners of yours. Okay, but you guys want to provide the bromide of the single pane of glass. Okay, I'm hearing open and agnostic. Okay, that's a differentiator. Security, you're in a good position to make an argument that you're in a good position to make things secure. You get the network and so forth. High performance, again, network and cost-effective. Everybody's going to make that argument relative to having multiple stovepipes, but that's part of your story as well. So the question, why Cisco? What's the key differentiator and what gives you confidence that you can really help win in this marketplace? So, our core companies is around networking and security. Whether it's cloud-based security, on-prem security, it's uniform. And from a security perspective, we have a universal architecture, whether it's the endpoint, the edge, the cloud. They're all sharing information and intelligence. That's really important instead of having bespoke products. These products and solutions need to communicate with each other. So if someone's sick in one area, we're informing the other one. So threat intelligence and network intelligence is huge. And then more importantly, as we're working with Google, Microsoft, and Amazon, we have on-prem solutions as well. So as customers are going on their multi-cloud journey and eventually the workload will transition, you have the same management experience, the security experience. So Anthos was a recent announcement, AWS as well, where you can run on-prem Kubernetes and you can take the same workload and move it to AWS or GCP. But the management model and the control plane model, they're all extremely similar and you have to learn anything new from a trading perspective. But okay, but I used to term agnostic, oh no, you did agnostic, I said open. But so you don't care if it's Anthos or VMware or OpenShift, you don't care. Don't care. And architecturally, how is it that you can successfully not care? Because the underlying fundamental principles is you can load any workload you want with this bare metal virtualized or Kubernetes-based containers. They all need the same. For example, everyone needs bread and water. It's just not different. So why should you be able to discriminate against a workload or OpenShift or you're already having a Cloud Foundry, for example. The same model, all applications still need security, visibility, networking, and management, but they should not be different across all Clouds. And this traditionally what you're seeing from the other vendors in the market, they're very unique to their stove pipe. And we want to break down those stove pipes across the board. So what happened, what work would you have? All right, Clay. So Dave, talk a little bit about the automation that Cisco's delivering to help enable this because there's skillset challenges, there's just the scale of these environments are more than humans alone can take care of. So how does that automation, I know you're heavily involved in the CXPs of Cisco. How does that all tie together? So working on a lot of automation projects with a large enterprises and SPs of, you see Rakuten being fairly prominent in the show. But more importantly, we understand not everyone's building a Greenfield environment, not everything is purely public cloud. We have to deal with Brownfield. We have to deal with third party ecosystem partners. You can't have a vertically tight single vendor solution. So again, to your point, it's completely open. Then we have frameworks. So meaning you have orchestrators that can talk down to the device through programmatic interfaces. That's why we see DevNet surrounding us. But then more importantly, we're looking at services that have workflows that could span on-prem, off-prem, third party. It doesn't really matter. And we stitch a lot of those workflows southbound, but more importantly, northbound to security and ITSM systems. So those frameworks are coming into life, whether you're a technical cloud provider or you're a large enterprise and they slowly fall into those workflows as they become more multi-domain. You saw David Gakler the other day talking about SD-WAN, ACI, and campus-wide and wireless. These domains are coming together and that's what we're driving more a lot of the automation work. So automation is a linchpin. To what business outcome? Ultimately, what are customers trying to achieve through automation? One is how to, there's this couple of things. Meantime to value. So if you're a service provider, to your internal customers or external, time to value and speed and agility are key. The other ones are on meantime to repair and meantime to detect. If I can shorten the time to detect and shorten time to react, then I can take proactive and preemptive action in situations that may happen. So time to value is really, really important. Cost is a play, obviously, because when you have more and more machines doing your work, your optics will come down, but it's really not purely a cost play. Agility and speed are really driving automation just at scale as we're working with folks like Rakuten and others. What do you see, Dave, as the big challenges of achieving automation when customers, first of all, I was talking like 10, 15 years ago, they're afraid of automation, some still are. But they, I think, understand as part of a digital transformation, they got to automate. So what are the challenges that they're having and how are you helping them solve them? So typically what people have thought about automation has been more network-centric, but as we just discussed, multi-cloud, automation is extending all the way into the public cloud at the workload or at the function level, if you're running a Lambda, for example. And the more important, traditionally, customers have been leveraging Python scripts and things of that nature, but then there is a script that's out there, but they cannot scale. You need a model-driven framework, need model-driven telemetry to get insight. So I think the learning curve of customers moving to a model-driven mindset is extremely important, and it's not just about the network alone, it's also about the application. So that's where we're driving a lot of our frameworks and education and training, and talent's a big gap that we're helping with without training programs. Okay, so you're talking about insights. There's a lot of data. And you're saying, goes data is plentiful, insights aren't. So how do you get from data to insights? Is that where the machine intelligence comes in? Maybe you can explain that. So there's a combination. So machines can process much faster than humans can, but more importantly, somebody has to drive the 30 or 40 years of experience that Cisco has from our tech, our architects in CX, and our customers in the community that we're developing through DevNet. So taking trusted expertise from humans, from all that knowledge base, combining that with machine learning, so we get the best of both worlds. Because you need that experience, and that is driving insights so we can filter the signal from the noise, and then more importantly, how do you take that signal and then in an automated fashion, push that down to an intent-based architecture across the board, right? Dave, can you bring us inside a little bit your touch points into customers? I think in the old days, it was a CCIE, his job, his title, it was equipment that he or she would touch and today, talking about this multi-cloud and the automation, it's like, okay, it's very dispersed as to who owns it. Most of what I am managing is not something that's under their purview, so the touch points you have in the company and the relationship you have changed a lot in the last few five years or so. Absolutely, because the buying center is also changing because folks are getting more and more centric around the line of business and what the outcome we want to drive for their clients. So the cloud architecture teams that are being built, they're more horizontal now, you'll have a security person, an application, networking, operations, for example. And what we're actually, I think we're pioneering a lot of the enterprises as SPs is building these site reliability engineering teams, or SRE, which Google has obviously pioneered, and we're bringing those concepts and teams through a CX framework, through Setelcos and some of the high-end enterprises initially, and you'll see more around that over the coming months, how SRE jobs, if you go on LinkedIn, you'll probably see hundreds of them out there now. All right, one of the other things we've been watching is, Cisco has a very broad portfolio. This whole CX piece has to make sure that, from a customer standpoint, no matter where in the portfolio with Core, Edge, IoT, all these various devices, I should have a simplified experience today, which isn't necessarily my words, Cisco's legacy. So how do you make sure, is software a unifying factor inside the company and gives a little bit about those dynamics inside? Absolutely, so, we take a life-cycle approach. It's not one-and-done. From the time there's a concept where you want to build out a blueprint for your digital transformation journey, we have to make sure we walk the client through just preparation, planning, design, architecture, optimization, but then making sure they actually adopt and get the true value. So we're working with our customers to make sure they go around the entire life-cycle from end to end, from cradle to grave, and be able to constantly optimize. I mean, you're hearing the word continuous pretty much everywhere. It's kind of the fundamental of CICD. So we believe in a continuous life-cycle approach that we're walking our customers end to end to make sure, from the point of purchase, to the point of decommissioning, making sure they're getting the most value out of the solutions that they get from Cisco. All right, Dave, we'll give you last word. Cisco Live 2019, thoughts, takeaways? I think there's just amazing energy here and there's a lot more to come. Come down to the CX booth and we'll have to show you some more gadgets and kind of solutions where we're taking off forward customers. Great, David, thank you very much for coming to theCube. Pleasure, thank you. All right, 28,000 people in theCube, bringing it to you live. This is Dave Vellante with Stu Miniman, Lisa Martin is also in the house. We'll be right back from Cisco Live, San Diego, 2019 day three, you're watching theCube.