 Okay, welcome back everybody. This is VMworld 2013. This is theCUBE live in San Francisco, Moscone, South Lobby. This is theCUBE, our flagship program. We go out to the events, we track the signal from the noise. I'm John Furrier, I'm with Dave Vellante, wikibond.org. George Schlesman is here. Hi everybody. We are excited to have a new segment. Everybody's talking about software defined data center, and George Schlesman is the CEO and product architect of IO.io.com. Go check out the website. Start up in this space. George, welcome to theCUBE. Thank you. My first question is, how'd you get a two letter domain name? Yeah, I want that answer. John is a smooth when it comes to these things. Yeah, the entire strategy for me is I wanted the shortest possible email address you can have. So, for everyone watching, it's at geatio.com. I'm excited to send you an email, but no, it was a... It's going to start flooding your mailbox. Yeah, I know how people play that. That's what's G at IO.com. It's as short as you can get. You can run a URL shortener business. Damn, that's a good URL. I mean, seriously, how old is that URL? Did you buy it? It's an old URL. We actually purchased it from an old web hosting company that was in Texas, and so they had had it buried behind layers and layers and layers of email servers and things like that, and had a couple 10 or 15,000 email users that were on it, and we just kept kind of pestering them over time, and it all got a hold of itself. Good acquisition. It's a good name, and it's a good act. It was a good use of capital for a startup. It really is to me. From my perspective, it's one of the most important things we did from a marketing perspective to the company. Yeah, I got the cube handle on Twitter, but I had to actually wrangle it. You're technically not allowed to buy Twitter handles. Oh, I didn't know that. So, we know it's a game, but anyway, I want to get that out of the way. It's always been intrigued. First question I wanted to ask you when I saw you in person, but let's talk about what you guys are doing. You guys are one of the hottest companies right now in the space of data center, data center in the future of the data center. I mean, right now, I did a post, Mark Hawkins knows back in 2008, I wrote a post, the internet operating system in the data center where we talked to HP Labs, the sensor networks using big data, making it a holistic operation. So, that's all about the physical plants and the physical assets. Now all the actions, software defined data center with Nacir is selling for a billion dollars, all that stuff. Great, software versus this is great. But still, there's a power problem, there's a footprint problem. Every facility's guys gets the phone call, CIO managing the budget. That's your world, right? So, tell the folks what the solution is. And why you guys are so successful right now. So, what we've done, what we've endeavored to do, John, is we've endeavored to bring the same software-enabled optimization that you can do at virtually every other layer of the stack day, whether at the storage layer, or at the CPU layer, network layer, or even above into application, platform, service, et cetera. We've endeavored to bring that type of flexibility into the physical data center assets and give you the ability to manage those assets in a way that correlates those to the workloads utilizing them. So, data centers traditionally have been managed in a facilities-based approach. They've been construction-delivered. And we had a very strong view that that paradigm had to break if we were going to be able to take full advantage. We, meaning the global cloud movement, is going, if you want full benefit of the cloud, you have to have a data center that can react to the cloud the same way that the cloud can react to the data center. So, if you think of what Google or Amazon or Microsoft or others have been able to achieve by matching their data center infrastructure up with the application infrastructure that runs inside of it, the efficiencies that they get access to, we can deliver those same types of efficiencies at an enterprise scale rather than at a web scale. And how we do that is we carve the data center up into modular units that we manufacture. So, you can now break bulk with the data center. You can buy just the amount that you need at any given time. And then by standardizing that delivery footprint and manufacturing the standardized interface for the physical data center, we were then able to coordinate that through an operating system called the IOOS, which brings all of the attributes that VMware or a virtualization platform delivers at the server level, we now can deliver at the individual data center level as well. So, thin provisioning, thin slicing, real-time optimization, dynamic control, and I'll give you one example of a real world work or use case that a customer is running today is, during trading hours run the data center for resiliency and security, off trading hours run it for efficiency and cost. And we can do that through software rather than having to do that through physical design. So, dial up or dial down the resources that you need for the data center, the physical resources in a granular fashion. Right, and so what I kind of joke about is, so we're at an event here talking about software defined data center, but no one's talking about the data center, right? So, strangely enough, CPU storage, network devices, we do it extraordinarily, I think, good job today of managing and optimizing. Application layer we're getting better at, data we're still struggling with, how we think about managing data in an early smart way. But the data center they've done nothing with. Like we, the industry in general still goes to a facilities team or to a real estate guy to try to solve the data center requirement. And the data center represents for a given organization, enterprise scale organization, anywhere between 15 and 35% of their operating budget on IT side is the physical data center. Yeah, I mean, most people take us through that because also not only is that the wrong people talking about data center, real estate and facilities guys looking for space, they're not tech guys, right? So, and it increases the cycle of deployment, so just the people mostly have these problems when it's like they run out of room, right? So take us through what you guys run into all the time with your, where you guys are knocking it out of the park. Absolutely, so two advantages our technology platform brings for our customers. As a customer's going through the process of growing or maturing into a cloud construct or a cloud utility model for compute, the data center tends to very quickly highlight itself as a boat anchor to that process. That they have this huge enormous capital investment into the static resource, the static asset, that they can't modify and change on a real time basis, right? Whereas everything else in the stack is going to a pay per drink model or a utility based model or a commodity vanity free open architecture. The data center remains this very, very, you know, monolithic static component. And so what we typically run into with customers is it gets expressed as we ran out of space or it gets expressed as we ran out of power or it gets expressed by we don't need that asset anymore because it's usefulness has run out to us and what we do is we then work with the customer to say how do you rationalize that and how do you take a step back and instead of making the data center decision on an enterprise scale to make the data center decision on an application scale. So if you think about one of the real quantum shifts that have occurred as we went from mainframes to PC servers and now to cloud, at the mainframe, in the mainframe world you made the IT decision on an absolute enterprise scale for everything. Operating system, hardware, applications, all ran in one device for the whole enterprise, right? As we matured into a PC server world, then we could make the decisions for that part of the stack. We really had a bifurcation, right? That part of the stack became an application based decision making and the hardware and the data center still remained an enterprise decision. The cloud is now allowing us to mature to a place where we can actually make almost call by call decisions about how we access resources. But again, the data centers remain completely static through this process. And so the burden of the data centers of showing up in these kind of four basic attributes for most companies today. One, they take way too long to deliver. The cycle times don't match up to consumption times and IT cycles. The burden of the cost on a unit basis and an aggregate is going up year over year and there's no way to rationalize it at a fine nine enough scale to be able to be proactive. You can't do anything about it, right? Some of these assets, God forbid, some of these assets are booked on our customers balance sheets on 30 year depreciation cycles. Could you imagine if a Spark server was on a 30 year depreciation cycle, like what the outcome of that is for the world? I got my facts over here. Yeah, right, yeah, it's all good. We're still depreciating, it's useful. And so those are how they're expressed. There's also an underlying security problem with traditional data center environments and that has to do with the systems that are used, the software systems that are used to control physical data centers are building automation systems, like literally like they were built to control air conditioners in office buildings, not to control data center environments. And so they typically aren't part of Active Directory or LDAP or any type of single sign on service. They're not part of the information security stack for an enterprise. And strangely enough, if you shut down one of those devices, you can shut down everything in the data center. So you said 15 to 35% of the operating budget is the facilities, right? Correct. So how do you affect that? You're talking about the power build with the beginning there as well. There's two components. So there's the inside of, if you want to express it all as operating expense, there's a portion of it that's depreciation, right? Or capital expense that's depreciated over time. And then there's the operating component of itself, right? So that's energy consumption, typically. It's personnel, it's all of the components that roll up into it. We affect it significantly two ways. First of all is the stair step process of procuring data center. We break into much smaller chunks, right? So instead of having to spend a hundred million dollars on a new data center facility, you can in fact spend one, you can buy one module if that's all the capacity you need, which could be cut to expense from a hundred to a million dollars, let's say. The second is by using our software, you can be much smarter about how you use and consume that service. The underlying hardware is more efficient in its design and the software lets you run it more efficiently so you can see operating expense costs come down significantly, energy. We just published a report that was validated by Arizona Public Service at one of our hosting sites in Arizona, which is the public utility. A 44% reduction in operating energy costs and operating expense from traditional white wall to our modular approach. And that was third party validate, we published the report. Not our numbers, they're numbers, right? They used our numbers to do it. But at the day that 44%, this is an enormous operating advantage just on energy consumption that you can see. I mean, you can go in and drop in and scale out data centers without all those hassles and then make them integrate in. It's a total home run. Well, and John, the second point to this that I think is further to the cloud point what we're here at VMworld about is, how do you then scale that from your premise to a hosted premise, to a hybrid model, to a private model, to a public model? How do you start thinking about on application basis, managing security, managing transparency, managing access for an enterprise, how do you scale out? So what are your top customers doing? So with that point, what are your top customers? Give me the example of your top customers. Our largest global customer today is Goldman Sachs. Publicly disclosed, branding's on our website, et cetera. So I feel free to discuss their name and what they're doing. They've looked at our platform and at this point in time, are committing to taking every part of our product set. So from our eco product, which is a high efficiency product, to our operating system running on top of those products, to our core product running in our hosted sites, not even on their premise, but in a hybrid, private environment. So at the end of the day, they have elected to take every component of our product set and deploy it inside of their IT stack as a cornerstone foundation element. Talk about the company, you're an entrepreneur. I mean, it's hard, you've been very successful. I mean, talk about the journey and how it all got started. As a Zen master says, you know, we'll see as it all gets out of hand. It's looking good off the tee. Yeah, it's looking good if it's the exact way. I'm in the fairway, so we'll leave it at that. Don't screw the approach. Yeah, right. So, you know, I've been an entrepreneur my entire career. My brother and I have worked together from our very first business, which we started in the bedroom of our, actually his bedroom, I have to admit, in when we were at university. First business in the application software development space, we were fortunate enough to sell to a public company. We then started our first foray into the data center space and really in the traditional wholesale retail collocation model in the early 2000s, had an opportunity to exit that business also to a public company. But really that was the moment where we said, these two worlds have to come together. Like, you can't have a facilities approach and you can't have an IT approach that's taking separate trajectories. And we said we had to bring those together. And in 2007, we started IO as a third startup. I mean, back to the folding tables and three guys around in a room. That's part. Yeah, no, it's funny. You're hungry, hustling, you're rubbing nickels together. No, you know what I mean, exactly. Scrappiness is core to our DNA and what we do because that entrepreneur being scrappy for your customers is a powerful thing and being smart and thoughtful. And you guys self-funded, right? Yeah, so. You had your own dough into it and then you self-funded it. Yeah, we self-funded the business to start and then we brought institutional capital in since then. But you were clear the runway by then, right? On the product. Oh yeah. Or did you bring in the capital? We brought in capital. First capital raise we did was in 2009, so. So two years in. Two years into the business before we brought in third-party institutional capital at that point in time. We're approaching, we're north of 400 employees today. We're about north of 500 customers globally. We have operations in Singapore where we have a hosting site for our technology, Phoenix, Ohio, New Jersey, and we'll have a site coming online in the UK next year. So that's where we can host our technology for customers. Then we have modules at this point deployed across the country and internationally on behalf of customers. How does the outlook look like? Obviously it's a trend. I mean, people are busting out. They still need data centers. They need data centers everywhere. So how is the outlook for the business? What do you see? Look, at the end of the day, we're going to be more applications tomorrow than are today. And I think that how those applications get deployed is getting smarter and enterprises who are primary focus are enterprise customers. So I'm not that we don't have web-scale customers, we don't have government, but our core focus are the big enterprise-scale customer. There's a few trends that these folks are focused on what I think are driving everything. Lock-in, no lock-in. They will not survive lock-in anymore. They will not accept it. Vanity-free. They do not want to pay taxes for brands, right? I mean, there's these underlying trends that they're clearly maturing in the space. And I think that anyone that can show up with a smart solution that's innovative and has a roadmap for progressively getting better is going to win. And I think that we're fortunate to be positionable. There's a lot of carnage in the data center expansion. One, that's getting the space. That's, you know, a facility's guy, real estate guy. That's kind of slow and not the right expertise. Two, putting it all together, like I was saying, this operating system model, it's basically an engineered operating environment. Now, every component has to work together. So I want to know more about like that. We're going to want to flash, so take flash. What are the inner engine components? Share with you your vision of that operating environment. And the benefits, obviously. So the vision for the data center from my perspective is this has to be commoditized. The data center infrastructure layer where we play all the way up through the enterprise infrastructure stacks of CPU network storage has to be commoditized. There is no way that we can survive at scale without commoditization at that layer. Period in the story. The number of components in that layer have to be minimized. And I mean the physical component count has to come down. I believe the only moving component that can be in the CPU storage network data center layer how we view the world can be fans. That's it. Discs aren't going to be spinning. Chillers aren't going to be spinning. Generators aren't going to be spinning. UPSs aren't going to be spinning. All of that over time has to go away. So we just heard our previous guests talk about specialized hardware and exact opposite story. I mean, I was surprised to hear it. You're like saying no way, won't survive, won't scale. The problem is it doesn't, customers don't want custom because custom equals expensive, right? And every turn, custom equals someone's proprietary lock-in that they want to hold your feet to the fire on. I see that if you look at where anyone who's achieved scale in data center, look what they look like. Facebook, open compute, Google, white label. They built their own, right? They built their own. And so we describe it two ways. We bring web scale to any scale. So if you want web scale and a web scale approach to infrastructure at any scale, we can do that. And the second thing is the future is that module, the data center is the computer. Yeah, remember the old DMZ era, it's the DMZ and it's the extra data and all that stuff. You guys can literally drop in a hybrid cloud tomorrow. Yeah, private hybrid. Absolutely tomorrow. Just drop in 90 days delivery. 90 day delivering, fully contained and engineered. And a serialized bill of, did I get that right? Yes. And a serialized bill of materials from a support maintenance warranty perspective, which is important for a large enterprise. Like you can't just drop in something for order of their own gear or you just decide for them. We give them the optionality. They can, we'll provide it as a service, either IS, pass or Dcast, or they can take it as an on-premise owned asset or we'll host an owned asset in our site too. What's the biggest thing people don't understand about your business out there? We know, Dave and I were talking earlier about it. There's always naysayers and the big booming companies like yours that are growing. There's always people who don't get it. I don't get it. Yeah. I think- What's the big thing that people don't get? I think what people don't get is that the data center isn't real estate. And I think that there's an entour, when you say software to find a center here, no one thinks construction companies are shovels and you know what I'm saying, they don't. Power and cooling, yeah. Because they're right. These are the people who use data centers and data centers are what happens inside of them. And I think the biggest misconception to the spaces that somehow a building is going to be relevant to the cloud in the future just isn't the case. How your operationalizing cloud infrastructure is going to be relevant and that should be outdoors or indoors. It could be at any type of location. The second thing is the software integration component I think is discounted and shouldn't be. And I would say discounted is just not well understood. We have built a platform that allows any advanced orchestration engine from a service mesh who's doing some incredible things in managing policy-based organization at top level to cloud foundry and Paul Moritz and the team out there are doing some extraordinary things with open components of that, open stack, other things, all of these things in the stack are building a construct where you should be able to dynamically control infrastructure second by second, call by call. And that's from the data center all the way to the top. And our software platform allows what VMware did for the server we do for the data center itself. And between those two things, once you decide the data center is technology and it should be optimized by software you understand what we do. Do you do it all in to ends? Do you do the assembly, construction and integration and software? One of the things I'm most proud of of our company and at personal level is we've created 100 plus manufacturing jobs in the United States this year. We've built and manufactured our modules in a factory in Chandler, which is an extraordinary site. We do tours there if anyone wants to come visit. And Arizona, in Arizona, Chandler, Arizona. We, I tweeted a picture earlier this year of our modules being unloaded off a ship in Singapore. Right, there's nothing cooler in the world than seeing something made in the United States shipped to Asia and unloaded. How many other companies in this venue can talk about how many of their products were built here in the US and shipped to Asia, right? I mean, it's exactly the opposite in every case. So, that's one of the things we're most proud of. So we have all of the engineering resources internal. We have an applied analytics team that takes the data we collect from all of our customers. We have about 20 plus billion rows of data we've collected in the metadata from the operating of our technology and customer sites that we do deep analytics against and then help our customers get better and help our product teams get better at engineering. How fast are you guys growing right now? So you're adding, you're actually hiring, right? Yeah, so two years ago we were less than 100 people and we're 400 plus today. We're hiring roughly a person a day right now from a month. What's the kind of people you're looking to hire right now? Engineers and sales people. Anyone who is an engineer, and the neat thing about our engineering effort is it's from hardcore mechanical electrical thermal all the way up through software. So we really have this completely integrated with that cloud engineering. You need all kinds of mechanical engineers, all kinds of systems engineers, software, integration. Materials, applied robotics, I mean all these sorts of things. George, really great story. I'm really glad you come on theCUBE. We're selling Dave. We'd love to have startups on but we founder-led companies are always our favorite because you start, you scrap together, you put it together, you fund it, you build it. Then you raise money to grow and scale up into cruising altitude. I appreciate that, John. But look at Michael Dell, he's going private. He's going to have three-quarters of Dell. He's going to take it back. And so I predict we're moving to a trend of founder-led startups. And you can see more and more of that. You only go to VCs when you need the cash, right? That's your care. That's your care. Of course, I'm biased. No, no, I think you're right. I'm the biggest, I'm one of the... Ownership, I think, breeds every responsible behavior we want. From leadership to caring. Like I said, I care. I get up every day and I care about my team. I see a bunch of them saying, I care about their 2,000 indirect employees and family members, right? Well, congratulations. Thanks for coming on theCUBE. We are here live in San Francisco. Go to io.com, check them out. Changing the data center for the better. Obviously the data center doesn't exist without one. I mean, you can't have a software-defined data center without an actual data center. And we know that everyone has power and cooling, huge issues in the data center, massive growth still at an issue. So facilities are exploding and they need to be monitored and managed. So, go to io. This is SiliconANGLE theCUBE with Wikibon de Vellante. I'm John Furrier, we'll be right back with our next guest after this short break.