 Live from London, England, it's theCUBE, covering .NEXT Conference Europe 2018. Brought to you by Nutanix. Welcome back to theCUBE at Nutanix .NEXT 2018 in London with you, Piskar. I'm Stu Miniman, happy to welcome back to the program, multi-time guest. Satyam Vagani is the Vice President, NGM of IoT and AI at Nutanix. Satyam, great to see you. And likewise too. Thanks for having me back. So, IoT and AI, two of the fastest moving spaces in IT today. Remember, we had you on the program a little over a year ago to talk about Project Sherlock. And that was where Nutanix was starting with IoT and remember asking you, Nutanix and IoT come on, I don't understand, Nutanix for the most part, it's like two products at that point and where they're going. The password today, it's now called XI-I-O-T. That's X-I and then I-O-T because Nutanix makes the product simple, but not necessarily the name. But Satyam, help bring us up to speed as to what your team's been working on and what's the state of the product today? All right. So, like you said, we started way back around a year and a half ago from now. And we were working on Project Sherlock. And the idea was the fact that a lot of analysts were projecting that a lot of enterprise data is going to be produced at the edge. And it was equally important to process it at the edge for many different reasons. Autonomy, security, cost, compliance. And so that was the genesis. And we thought we were very well-suited to do it because this was yet another problem where you needed to provide a very elegant system, a very well-contained system, just like what HCI is for your primary data center in an extremely remote and extremely hostile environment. And so that was why we thought we should take a crack at it. And along the way, of course, then our ambitions brought into being a multi-cloud company. And that fit in very well as well because IoT is never an edge-only problem or a cloud-only problem. Every IoT app kind of spans the edge in the cloud. And this was a perfect way to showcase the multi-cloud data plane, multi-cloud control plane capability. Yeah, no, I love that. It was one of my rants this year is, you know, oh, you know, edge kills the cloud. I'm like, no, it doesn't kill the cloud. Edge actually will help increase a lot of the stuff in the cloud. It just means that it all won't be in this mystical the cloud in the central, which we know everything's being built really as it's distributed architecture today. So there's a lot of nuance and understanding, you know, how some of those pieces together. I was excited to hear, if I remember right, you have a customer talking publicly about using this today. So there's nothing better than hearing, you know, a real-world example. So maybe help walk us through that briefly. And that's the most fulfilling part about it. So, you know, that's a plan for tomorrow's keynote. And, you know, it's very fulfilling because when we started IoT, you know, one of the other concerns other than the why Nutanix was why IoT, it's, you know, not a mature space. Nobody quite knows what to do about it. And that was the point is that we thought we had an opinion about the edge being us kind of key piece of IoT. The edge plus cloud conversions being a key piece of IoT, the centralized management being a key piece of IoT. And then we were able to validate it, not just with, you know, POCs, but, you know, with people who have put us in production in very fantastic and remarkable use cases. So, yeah, you know, that was the journey. This one use case was around smart retail, but it kind of embedded all the elements that we kind of hold dear to our heart is, you know, can we instantiate AI inferencing, complex AI inferencing logic at the edge? Can we instantiate standardized applications at the edge? Can we do interesting data management between the edge and the cloud? So looking at this from a technical perspective, what does this look like? So if I look at this from, you know, from that technical perspective, I still see a data center, I still see the cloud, I still see data going back and forth. Yeah. What makes Xi IoT different? Different. So one is the focus on edge computing. So, you know, a lot of IoT solutions were either made at a time when there was a kind of vertically integrated IoT application and that vendor said, okay, now let me get broader. So that's one category of, you know, IoT solutions out in the market. There were some IoT kind of offers, which were cloud first, right? So, you know, there's an IoT offering in the cloud potentially for doing consumer IoT projects. And now that offering wants to expand into enterprise. We said, let's go from the edge outwards because, you know, in the enterprise context, like I said, data processing, the amount of data, the volume is pretty overwhelming. So that's one difference, which is the richness of services that we provide at DH. The stack is pretty deep, but at the same time, pretty miniaturized to Dheeraj's point from the keynote in the morning, is these environments need to run in pretty compact form factors in terms of computing. So that's one difference. The other difference is the pipeline all the way to the cloud. We don't consider this as an edge-only problem. And to that end, we normally do a pipeline to the cloud, but we allow the customers to have a choice of cloud. So we don't dictate a choice of cloud just because we are providing a solution to the edge. And another key difference is the ease of use both for deployment and operations of the edge device itself. So you know, think about deploying this thing or a thousand stores. We made it a zero-touch provisioning process. So the only requirement to deploy the Xi edge is that you plug in the internet cable. And that is very core to the Nutanix philosophy, right? Simplicity, one-click simplicity. And the last thing is APIs. Is the programmer APIs that the whole system exposes is Apache-class APIs, open-source-class APIs, so that people who are already used to various programming frameworks can immediately jump on this. Yeah, I mean, you bring up one of the, what we saw in our research, one of the biggest hurdles for this. Go back to when we first looked at industrial internet. We actually did some research with GE, and it's the OT really doesn't play with IT. You've got very specific gear, and it plays certain ways, and it doesn't talk, and it doesn't have APIs, and it doesn't have networking, and it's all going to have centers and connectivity and things like that. And just these worlds just don't talk today. So is it, it sounds like this is a more IT, friendly solution, how do you help bridge that gap? That's a great observation. So first of all, I would say, indeed we are coming IT inwards, right? IT outwards into OT. But at the same time, you know, the only way to make OT appreciate such a solution is to show them a path that look, you can adopt Xi IoT without causing disruption to your machine critical setup, OT setup that you already have. So we've put in a lot of thoughts around how we can source data from OT systems without having a conversation about throwing out, ripping and replacing every OT data gathering point and device that they have. So that was one thing. The other part is if you can provide them some extra added benefits, if that OT person was to do some of their infrastructure on top of this kind of IT OT convert system. And so to that end, we think there are some specific security benefits, some resource management benefits, some user management benefits that we can provide in this new age that the OT guys would appreciate as well. And so it was about having something for the OT guys to appreciate so that there is some buy-in as opposed to dictating that you got to do it. So one of the things I observe in that industrial world is there's quite a lot of developers in that space as well. They're actively developing while gathering data while figuring out IoT. So how do you let them, you know, let those developers work on that platform as well? For instance, do you use carbon on that platform, stuff like that? Right, so you know, we have a bunch of services on the platform, essentially it's like a pass, but for the edge. So there's container services, functions, there's some data services, AI services. But the developer point is very interesting. So in fact, as we speak, we are going through that journey as a company, right? How can we be a more developer-centric company? You know, it involves literally running a seven-hour XI IoT lab yesterday, and a lot of people showed up and they stayed throughout the day. So I think it's an awareness thing. It looks like at least from an interest point of view, once people see the platform, the APIs, there is true interest. And so now it is up to Nutanix to have enough events, have enough kind of awareness campaigns to make sure that this work spreads out. Unfortunately, I don't have a silver bullet for this, but it's literally going to be a work in progress. I think next year, same time, maybe when we talk, probably we should talk about whether we make progress on this or not. Speaking of areas that don't have silver bullets, you mentioned security. A lot of concern around, there's already been breaches and hacks and things like that. Security, as we know, the surface area goes up in order of magnitude or more. How does Nutanix look at the security aspect? So there's various things, but I'll give you one example. Concrete example of what we did that is making a dramatic change, moving the needle, so to speak, is if you can always argue that your platform is secure, but at the end of it, how do you prove it? And so one proof point we have is from a security point of view, the Xi IoT edge is locked down. Even an administrator cannot lock it. And so the idea was that if you ever let an administrator have a username and password to an edge device, that is bound to be a point of compromise, no matter how secure you are. And so the only way to eliminate that is to just eliminate the need to have a username and password to an edge device. And so those are some of the things that we kind of thought and that's actually tightly ingrained in how the system is designed. Because you can't touch the edge. You can deploy all the applications you want, but you can't touch it. You can't touch the, you know, we provide containers as a service to the edge, but you don't configure container isolation on the edge. The system does it for you. So these are some of the things that, you know, the more automation we do, the more we remove humans from some of these very mission-critic, security-critical points, the better off we are in terms of, you know, reduced chances of hacks. Okay, so you've got a customer up on stage tomorrow. Is Xi IoT, is that now GA today? All right, sorry, I forgot to make it. It is, indeed, it's GA. Okay. And the press release went out today. Okay, you know, give us, what should we be looking going forward? We understand we're still as an industry early on this journey. Your team's been working on it for a while, but what should we be looking for? What are some of the key things down the road that you're excited for? From Nutanix, or industry? Yeah, well, I'll take both. Okay. Give me two bars. Well, anyway, let's see. First from Nutanix. So, you know, we want to, IoT is not a single vendor problem, right? So, as much as we want, we want to make this a platform that is, you know, attractive to OT folks, that is attractive to IoT vendors who have been creating these very vertical specific IoT apps. We want to prove to them that this is a way better platform for them to deploy their apps at industry scale. We want to appeal to AI guys, data scientists who are going to create interesting applications around data processing. So, you know, some of the next few steps is to provide interesting features and functions in the system which appeal to all these demographics. From a GTM point of view, go to market point of view, we want to make sure that we get some partnerships, right? Because again, this is not just a technology problem. So, those are some of the steps you are going to see. You are probably going to see more we are open to, you know, to bring your own cloud philosophy for our IoT platform. Out of the box, we support Azure, AWS, Google, and our private cloud, if you bring that in. And maybe we'll expand that portfolio because there are other cloud providers, especially if you're looking at regional markets like APAC and AMIA and so on. So, that's some of the things. And then, last but not the least, you are going to see more and more investment in AI because, you know, there's obviously a lot of talk about AI. But, and it's very easy to do a proof of concept, or well, kind of easy. But it's very difficult to deploy that proof of concept at an industrial scale. And that is a problem we want to really, really solve very, very well. So, you'll see a whole bunch of investments, features, announcements around it. All right. So, I think we're going to have to leave it there, Satya. Pleasure is always to catch up with you. Congrats on the progress and look forward to hearing from your customers. Thanks very much, Stu. Thanks, Joe. Thanks so much. Hi, I'm Stu Miniman. I'll be back with more coverage here from Nutanix.NEXT 2018 in London, England. You're watching The Cube. Hi, I'm Stu Miniman. I've been an analyst with Wikibon and a co-host of The Cube since 2010. Before that, I've worked in the tech industry for many years at a number of different companies. My background really is in networking, virtualization, cloud computing, since the early days. I really love the intersections of some of the technology.