 Live from Miami Beach, Florida, extracting the signal from the noise. It's theCUBE, covering.nextconference. Brought to you by Nutanix. Now your host, Dave Vellante and Stu Miniman. Welcome back to Miami Beach, everybody. This is Dave Vellante. Stu Miniman, D-Rush Pandey is here as the CEO of Nutanix Next Conference. First conference ever, you must be thrilled. Almost a thousand people here, congratulations. Thank you, it's exhilarating to say the least. And then being the first is always special. And Miami, the hotel, the atmosphere, the perfect demos, the vision. Monkey dance. Yeah, the monkey dance, you know, all that. I think it all just settled down well, so. We're struck when the conference got started, the keynote got started, everybody expected to see you. It was your first customers up on stage. And we now know they weren't using teleprompters because they were fumbling their lines a little bit, which is beautiful, perfect for engineers. Who came up with that idea? It was genius. I think it was a group effort. Obviously Howard Ting, our SVP of marketing, is also a genius when it comes to really talking about messaging and either showman. Just like every good company needs one, you know? And the best part is that he's also had product management background in the past, so he knows how to weave product with marketing. Talking about the three H's. Hungry, honest, and humble. And I think it's people who know you, know you as a humble individual. Maybe you hire showman, but you yourself. Where'd that come from? What's that mean to you and your customers? Also people confuse the way we deal with competitors. That arrogance as, oh, they're not humble. And I say, you know, within the family of employees, customers, and partners, we are extremely subservient. I mean, borderline subservient, actually. With the competition, we are a bit different because at the end of the day, we don't have an overlap of interest with the competition. So we are a little bit schizophrenic, you can call that, you know, when it comes to dealing with competitors, dealing with people who really matter. And I think hunger and honesty are just, I think it's ingrained in the DNA of this company now. And I wouldn't say it's me or it's just everybody that came together and said, look, if you don't build honest products, you know, you'll actually get killed along the way. And that paranoia actually has kept us honest. You know, when we hired Stu in 2010 out of EMC, we said, Stu, pay attention to what Google and Amazon are doing and then project that onto the enterprise, because whatever they're doing, the enterprise is going to be doing in five years. You guys must have had similar conversations and actually put that into products. Can you take us sort of back to the beginning? Absolutely, I mean, if you go back 15 years, actually, not just the founding of Nutanix, but 15 years when Linux was being tested in the enterprise, was basically the web skill guys who really tested Linux to the health. When it came to IDE drives that became SATA drives eventually, it was basically the web skill guys doing it. MySQL, the web skill guys did it first. All of- Yeah, the LAMP stack. All of Hadoop and the LAMP stack actually came from the web skill guys. Flash, actually. The first mass scale use of flash was actually Google and Facebook. Yeah, Fusion.io. And then Google was using flash for auto-complete and all that stuff. So, and from there on, everything is actually seeping the enterprise. So, the architecture, CPU and the enterprise were not anything new for us. We knew it was going to happen. The big challenge was legacy applications. What do you do about them? Because you can't just throw them away. One of the luxuries of web-scale companies is that they just write applications bottom up for the first time. And they don't have to worry about carrying the legacy along. So I think the big thing that the design point we had was what do we do about legacy? And we very religiously said, look, we've got to carry the legacy along. Yeah, I wonder if I can unpack that a little bit. I think back to the early days of server virtualization. In many ways, server virtualization helped me stick with my old applications. I mean, I think back in 2002, one of the earliest use cases I knew of was, hey, Windows NT, the server's going to go end of life, end of support, I can stick that in a VM, I can leave it there forever. And I can change the underlying hardware. So server didn't matter anymore. The big air gap we have today is how do I get to more modern applications? What's going to enable it? Things like platform as a service, containerization all kind of help that. How do you guys help? As I said, there's the old, there's the new, transition stuff. What's Nutanix's role in that transition? Yeah, so obviously the architecture is the same both ways. If you look at the scale up apps or the old client server apps, and if you look at let's say MongoDB or Cassandra or any of these next generation apps, we actually have a common substrate, the distributed storage fabric, the app mobility fabric, they're all going to be able to work across both these islands. And the next generation apps actually still need the 20 years of tax that Oracle paid with resiliency, availability, disaster proofing, all that stuff. How do you do data guard? Which is Oracle's flagship disaster recovery product with Cassandra or with MongoDB? How do you do strict consistency? How do you do geographical replication, synchronous replication, backups, all those things that those taxes have to be paid. And we just pay that automatically. Basically we've paid all that stuff in the last five years. So what partnerships are most important for that? How do you kind of build an ecosystem? We kind of say, IT today is moving from point products up to a platforms with extensible APIs and you can't do it along. I've seen some of the partners you've here, but is Nutanix the platform? Is, where's the platform? Where's the ecosystem? How's that fit? You know, I personally like the word platform but I also don't like it because the notion of platform has an element of hubris built into it. The fact that now you have arrived. It's the thing that really came together for everybody. And there's a right of passage to becoming a platform company. So if I were to say we are a platform today, I'd be remiss if I said that. But I think working really closely with the application guys, having all our people, all our employees understand apps with solution architects and sales engineers and support engineers and our developers to really understand apps and that we have done a pretty good job of. When I say that we've carried legacy along with us on this new architecture, it's because of our understanding of apps and not just apps with security, better than anybody else. So we have folks who really understand Oracle Deep and Microsoft Deep and Splunk Very Deep and so on. So our best friends are actually going to be the app guys because they are also the ones who actually struggle to get time to market going when they actually go to deploy something. Splunk guys they suck wind when it comes to time to value because it takes them forever to carve out a learn from EMC and get some server capacity and then realize it's not a capacity issues, a performance issue because it's Splunk indexes won't work and so on. The exact same problem that virtual desktop guys had are the problems that Splunk guys have, the problems that Oracle Database guys have and so on. They all would like to really figure out how to embrace infrastructure that's invisible. So the gardener guy Ray Poquette talked about mode one and mode two. Stu, you mentioned IDC platform two, platform three. Are you like mode one and a half? Are you a half way house? Are you driving toward mode two? I wonder if you could sort of clarify that. I think I would look at us as mode one point five because we're trying to really take the legacy along with us, the legacy apps, although the architecture is mode two. So, and this is the only way the baton would have passed gracefully otherwise enterprises would have given up. You saw what happened to all the open stack companies. They're nowhere, Nebula, and this and that but they didn't think through legacy. Legacy was a very important part of building any large business actually. So I would say that from the app point of view, we are at one point five. The architecture point of view, we are at two and the two is dragging the one along to get to one point five. So it's like Vinod Kosil said, if sun was under pressure to become a VT 100, you know, lookalike for a deck vac. So you have that similar pressure. How do you fight it off? I think it's going to be a combination of packaging and distribution beyond just technology. And I'll tell you what I mean by that. So DevOps consumes products differently. You want to buy a commodity off the shelf and put software on top. So quite a few of these guys would actually be comfortable buying software from us. And it's going to price that scale with, you know, sort of, you know, let's talk about how many sockets we can actually buy the entitlement form, so on and so forth. As opposed to folks who want much stricter SLAs who will buy appliances from us. So we need to have the buy-mortar way of distribution as well. It's going to be appliances for legacy apps and software for next generation apps. And you don't want to over rotate to the future. You end up like Nebula. At the same time, you've got to focus on the skate to the puck, as Vinod said. Yeah, and I think, you know, companies that get scared of, you know, what the market forces are actually asking to other ones that die. Like, you know, son was scared of X86. And they even squashed it. They're saying, you know, Intel is a toy and Microsoft Windows is a toy. And finally, by the time they came around to really embracing X86, they chose AMD as a partner as opposed to Intel as a partner. So it was too little, too late to really get opt around to run Solaris, I think. We understand that early enough to know that consumption models are things that we can't take sides on. Appliance is not the only way in which people consume our technology. And as long as we have that, we'll actually be able to, which is why Amazon, we don't have an appliance in Amazon. We don't have an appliance in Azure. Form factors are one thing that we have to be extremely flexible on. And that's the way we actually straddle both the world. Because the problems of consistency, availability, reliability, disaster-proofing, network partitioning, all those problems are similar. Whether you look at the old world or the new world. I mean, a lot of people start feeling a little excited about, oh, it's all going to end up in the app. The app is going to have all this value. I'm like, how many .NET developers and Java developers even understand the cap theorem, which is this trade-off between consistency, availability, and network partitioning? How many of these, most often, they actually want to go the other direction. Having less detail, abstract, and abstract, and abstract, and you know, I want to write a single line of code that instantiates a shopping cart. That's what they want to think about. As opposed to, hey, can I really? So the layers below have to really go and do that, as opposed to the application itself. All right, so yesterday when you talked to the analysts and the press and everything, you said, our goal is to abstract infrastructure, deliver invisible infrastructure, and elevate IT. The challenge is if you truly make it invisible, how do you prove that you're valuable? How do you just say, oh, that's stuff in the corner, and oh wait, there's four other companies that also said, hey, I'm going to be invisible. We've heard SolidFire, EMC, lots of others say, oh, we understand that storage especially has been hard. So how do you make the people that buy and people that use your products heroes and know that you guys are very important? Yeah, I think the way we're making them heroes of the future, I mean, take a step back. If Kodak engineers had seen digital photography coming, they would actually have embraced that sooner. They just were too much in denial of digital photography. I think we have to let storage engineers realize, and even a virtualization engineers realize that, keeping the lights on is not good enough, because the public cloud is a real enemy. If you were to look at jobs, that's it, because Amazon might not actually employ you. So if you want to really make your job successful, then you need to make the private cloud as efficient as the public itself. So don't worry about the fact that you have to give up on some of the legacy architecture. Worry about where the line of business is pulling you, and the line of business has a lot of people who are rebels. You know, VP of sales rebelled and said, we're going to use Salesforce.com. VP is a marketing rebel, and they said they're going to use Marketo and Eloqua. VP of engineering are using Atlassian and so on. So the trend is there, and I think for all the custom stuff they're doing, it's going to be infrastructure to service, unless they really go and make the private cloud as efficient. I love one of the quotes from one of your customers during the keynote was, the days of managing raid groups are gone. So there's a little bit of fear out there that, oh wait, are you killing storage jobs? Does storage, you know, knowledge go away? But we don't think that storage is critically important. You're just kind of elevating them, allowing them to focus on the business even more. Yeah, I mean, look, cobalt engineers who actually were able to re-skill themselves still have great jobs. Well, that's the key though. I mean, if you're an expert at lung management, you better not keep honing those skills. You better find some other things to do. You know, and that's why we have a great education department that's actually talking about re-skilling a lot of IT as well. We're not just talking about technology. You're saying, let's figure out through Udacity and through our own programs, how people can re-skill themselves to learning the next generation technology as well. So a lot of talk about IPO at this event, I know there's not too much you can say, but it's clearly something that you've put out there. You're one of your VCs Ravi was talking about the doom loop. I love that doom loop concept, which what he's talking about is, I call it the vortex of incremental investment. When you go public, if you go public, there's that pressure to get sucked into the quarterly doom loop, the 90 day shot clock. Now others have avoided it. I mean, you certainly see the folks at ServiceNow and Splunk and Tableau continue to innovate. Do you think about that? And how do you think about that? Yeah, I think we've laid the right foundation of things like remedy recognition and things like that, where we're not trying to borrow from the future. You know, there's a word that people use for companies that absolutely must sell a lot to meet a quarterly number. And then the others who just have deferred so much that the waterfall from previous quarters is enough to meet the numbers. And many appliance companies were on the former path where they kept borrowing from the future, not giving enough to deferred revenue but recognizing everything upfront. They have to sell a lot of boxes every quarter, every quarter. And people even call that crystal meth. Once you're in crystal meth, you can't get off of it. But if you have laid the right foundation of deferred revenue, where you're actually giving more to the future than actually borrowing from the future, then you do set the right foundation where by the time your quarter begins, you're enough coming from previous quarters, the waterfall from previous quarters coming in. And we've been a pretty disciplined company with that. As and when our books become public, you'll see that discipline from the mechanics aspect. So, Dheeraj, one of the challenges in the valley is there's always that kind of next thing. You've got now 1100, what you call the Newtons out there. How do you keep hiring? There's still enough good people to move what you're doing. And how do you build a company to last, not just to kind of hit the next thing? I mean, I said this in my short little keynote as well. At the end of the day, we have to think on behalf of the employee. And if we basically just said, what are they trying to achieve? Will they are asserting themselves for a promotion or a career advancement or a compensation? At the end of the day, we just have to think from their point of view. And the moment we do that, it disarms everything. It just disarms the whole relationship. They're like, oh yeah, I think we know where they're coming from. And some of these people actually need coaching because they might not know that they're not as good enough and yet their others were really, really, really good. That might be underpaid. So, we have to really look at both sides of the coin. And obviously, we have to look beyond Silicon Valley, which is a company we have done. Laid the foundation for Seattle and Durham and Amsterdam and Bangalore as places where we'll actually hire talent as well because these are the places where, once the brand is built, the talent actually says I'm willing to work for this company. That's what happened in Durham. We're getting a lot of really good people from NetApp and other places. Amsterdam is great support engineers being hired from all around the best valley companies. And Bangalore is another place where we're actually hiring as well. So, I think in the spirit of distributed systems, building a distributed system, which is not a single point of failure in the valley itself. And within the valley, we have to figure out how to really take care of our folks and make them realize that they have a career because once they know that this is not just the end of all things Nutanix and it's only the beginning, look at the most iconic companies, what they went public at and what they are today, they are 100x bigger. And obviously people are like, well, I want to make sure that I'm also relevant when there's so many more people. And we are constantly thinking about those things and it's not an easy problem. I can't say that we have a solution, but if you look at culture at 1100, we're actually, and you can go and take a look at Glassdoor, for example, it tells you about culture of the company. We're actually doing better than many companies were smaller than us. Now that being said, I'm a very paranoid person so I know that the 1200 guy can actually save the opposite of that. So we need to constantly monitor, measure, understand with surveys and make it local. Culture is never global. We have to know that the culture of China is different than culture from Singapore. Employee culture is different. The top-down mentality in some of these countries is different than the bottoms-up grassroots strength that we provide here in the valley. We put all that stuff together. So last year at VMworld, we had you on and it was coincided with the EVO rail announcement and you responded, hey, welcome to the party. It's confirmation of what we're doing. But there's been a little bit of what I call a urinary Olympics between VMware and Nutanix. Particularly, you guys decided not to OEM vSphere and VMware is kind of using that in their blog sort of to attack you. Can you clarify sort of that decision? Why not OEM vSphere? And what does that mean for your customers? Yeah, I think it's pretty clear to us that we don't want to be a reseller of a technology and simply because we didn't want to compete with the channel and the channel already makes, let's say 10 to 15 points selling vSphere and other such VMware software. And if we were to become yet another middleman, either we would have to bypass the channel or it would become more expensive for the end customer. So we were pretty firm about it. We didn't really want to go and do something that cannibalized the channel. And we are a 100% channel company. And the one thing that we obviously didn't want to do was to become the enemy of our biggest friend from the channel itself. So it's pretty crystal clear in that. And we went and told VMware that look, we are an app on top of you. And this Oracle runs on top of you and Microsoft runs on top of you. We don't force them to actually ship an OEM your product. Why should we as an app? I mean, that's what we keep telling people. Storage is just an app. I mean, why treat it any differently actually? So I think logically also it didn't make any sense if Oracle were not OEMing it, Microsoft were not OEMing it, and Splunk was not OEMing it, why shouldn't that be OEMing? Right, really didn't see any way to add value there. So why make it your core to your business model? Yeah, talk a little bit about Microsoft. So some interesting things. If I look at Acropolis, it can help customers get over to it. Of course, if you talk about from an application standpoint, Microsoft apps are one of the major workloads in a virtualized environment. How important is Microsoft to your plans? I look at the post Satya Microsoft as a Microsoft that's really thinking clearly on multiple fronts. As an example, they frown upon the strategy tax of the platform. They're like, look, no more we look at ourselves as a platform company and the office app has to work on Windows only and will never ship an iOS and so on. They've completely obliterated all that. And they have a clear thinking like startups do actually, where it's a hungry mentality and a humble mentality about the fact that there's no such thing as a platform. It's all about apps and it's all about customer delight. They also are a company that really has an inside out sort of understanding of the private cloud and the public cloud. But Azure is a pretty decent public cloud that's working at scale. So we as a company, look at them as curationally. They have a pretty good foothold in the private cloud at least with respect to the app itself. But they also have a pretty good foothold in the public cloud with Office 365 and other such things are doing there as well. So it's great to learn from them and actually enable this notion of hybrid cloud computing using Azure as an example. Now this is not to say that VMware software and the VMware stack cannot use Azure. We will make that work actually. VMware stack in the private cloud working with Azure in the public cloud. VMware software in the private cloud working with Amazon in the public cloud. All those are possibilities that we are constantly thinking about and making work as well. Yeah, well, we've said at the top, we see VMware slowing down certain initiatives. You guys are speeding them up. Paying attention to the customers. D Raj, really thanks very much for coming on theCUBE. Congratulations. Thanks for having us here. Great event to be at covering. All right, this is theCUBE. We'll be back right after this word.