 Pet here. You are the Senior Technical Director for Joyant, correct? That's right. What exactly does that mean? I'm a field-facing sales technical lead. I cover all of the Americas. When you get into a small company, you get fancy titles, but essentially the technical guy for all of the Americas. So you go out and you say, what's your problem? What's your problem? How can we help you? I think we can help you and go through those things. How was the conference so far for you? Have you been here all the time? I've never been here before. This is your first doll? This is my first doll event, yes. Is this your first storage event? No, EMC World. You did EMC World. That's right. Because you have some EMC legacy too? I do. So for about a year I worked at EMC prior to coming to Joyant, and then before that I was an EMC partner. So I lived very, very lively in the storage area. So far so good. I'm pretty impressed with a lot of the innovation that Dell's doing. They're releasing a lot of cool products and kind of keeping up with the market. And they're doing a lot with social media. That's been pretty impressive is a lot of the people that have been working to do a lot with social media. So I always like to see community involvement and getting people engaged. So with this, we were hearing a lot about the hands-on lab and the ability to come in and touch things, break things. I don't know about break things, but try. So it seems like that's becoming the growing trend with vendors, right? At the end of the day, customers want to get their hands on the gear and play with it. And when you're known for ease of use, like Dell has been known with their ecological acquisition, you need your customers to be able to touch and feel and smell the gear and play around with it. That makes sense. Have you gotten some of your hands-on? I have not. But I hope to get my hands-on some of the gear later on and play around with it. That's great. So we've been talking a lot about cloud stuff today. And private cloud versus public cloud, you kind of with Joyant work with the private cloud? Private and public. So Joyant started out about six years ago as a public cloud company. But what happened was about a year and a half to two years ago, Dell and Intel came to us and said, you have a pretty compelling platform that's different from VMware and a lot of the other offerings. So they said, why don't you productize that and actually go out and sell that to service providers and enable them to offer Amazon-like services at a competitive price. So we worked a year and a half and we are product G8 in April. April of this year? April of this year was our 6.0 release. We actually were selling an older release with Dell that we had productized about six months before that. So Dell DCS solutions. Yeah, data center solutions. Yeah, they're data center solutions. Or they're Dell cloud solutions, I think. I think it's Dell cloud. I don't know. I kind of get mixed up because some people use different terms. We'll look, right? So it's specific for web applications and they actually take our solution into enterprises and enable them to do things on our platform that they really can't do on other platforms like VMware. Okay. So that's going well. It's been going well. Being a Dell partner has been pretty exciting for us and then being able to actually have a channel to go into the field has been good for us. Were there any nervous feelings about becoming a Dell partner? You know, I don't think so. I think the nervousness was around, how do you work with an OEM and then how do you actually... We were a services company before. So switching gears and becoming a software company and selling to enterprises is difficult for somebody who's been such an engineering heavy organization. So Dell really helped us on the, you know, how do I go to market? How do I start executing? How do I support it and stand it up? So Dell does integrations and support as well for us. So for the solutions that they sell, we also do sell direct service providers, but Dell has been probably one of our biggest partners. It sounds like it's going fantastic. It is. It is. It's going really well. So this whole migration public to private or hybrid clouds, where do you think we're going to end up? So it's kind of interesting because I, you know, I'd like to describe a lot of the characteristics about what enterprises and companies in general like about cloud and where they put it really doesn't matter to me as much. And the reason why I say that is because when you walk into an enterprise or company in general or even dot coms like LinkedIn, Guild Group, those kind of companies who are our customers, you have, you find these companies that have traditional legacy enterprise workloads. And then you also have these newer organizations within that are doing innovation. They wanted to stand things up quickly. They want burst ability. They want scalability. Yes. We were talking a little bit yesterday, burst ability. Do you know that term burst ability? Yeah. It's not about bubbles. No. Really? It's not about bursting bubbles. I know it a little bit, but go ahead and delve into it. Yeah. So they're looking for, we can use elasticity as another term. Elastic cloud. No, I think that's taken. That is taken. But so people are looking for different characteristics where in the enterprise, traditionally, they're looking for infrastructure itself to provide the resiliency, the scalability. And then as people go on to newer types of languages like Ruby and Python, and they're using a lot of different technologies to develop quicker in the enterprise, they're looking for a different platform that will allow them to do that. And so with us, we have, again, burst ability. So you start up your website. Here's a good example. You start up your website. You have no idea what kind of traffic you're going to have. And if you're using traditional things like VMware where everything's just put in a little boxed up container, if you hit a spike in load, you have to call your service provider back or shut your machine down, add more CPUs to it, and stuff like that. What's been so attractive to companies like LinkedIn is that they might not know what that traffic's going to look like at any given time. And so we actually, our software allows them to burst and use all of the CPUs in a box. They can use all the IO capability. And then we still provide multi-tenancy so that they're not stomping all over their neighbors. We make sure everybody has their fair share. We have that ability to burst without re-architecting your application. I'm curious with LinkedIn after they did the very successful APO, did they get bursty that day? They did get bursty. We were really happy about it, too. The interesting thing about LinkedIn is that their core IP and this can kind of go back to the private versus public aspect of it is the engine that drives the database talking about the relationships and who knows who, that's actually all in their data centers and they kind of hold on to that. Everything on the outside of who's tweeting, who knows who, you know, who's status updates, all those kinds of things, they basically had this, I think they call it a light engineering team. And what they do is that they come up with all these additional features that they're going to use. They have no idea if they're going to be successful or not. So they deploy those things on our platform that tie back into their database that they're hosting on their own. So they have a core database which isn't based on yours. It's not based on ours. Sorry. Sorry for your ears. Yeah, sorry for your ears. But they have all these core, these emerging applications that sit on the outside. That's a perfect way to put it. And then another example is the Guild Group. So the Guild Group isn't it? Guild? Yeah. Oh, Guild or Guild? Guild. Yeah, so Guild is GILT. What did you tell them? I don't know. I made somebody feel bad about something at some point in time. So they have this kind of flash mob sale thing where they'll send an email blast out to all of the people who are signed up. And there will be a sale on a specific set of items. Okay. And it only lasts for a couple of hours. So their traffic is very bursty. Yeah. And so there are over a billion dollars in revenue and everything is hosted at joint. Wow. Yeah. And their traffic misses go like... It just goes straight up and straight down. How interesting. So they can scale down as easily as they scale up. They can. And so their cost is that fluid too. Yeah. Yeah. It is, right? It is. It does. It does. It's very fitting. So that's the exciting thing is that you see more and more enterprises starting to do. It's not just e-commerce and social media stuff, but you have an example of somebody using Cloud is Eli Lilly is using Amazon Web Services. Now we allow service providers to compete with Amazon Web Services, but it's an example of an enterprise using Cloud and bursting and elasticity and, you know, utility billing to actually do some cool things. So what they do is that they're actually doing pharmaceutical research inside of Amazon's Cloud and you would start to say, okay, there's security concerns, intellectual property. They've done things internally to protect themselves like they encrypt their data. But what they find is instead of using a supercomputer to do work, to do pharmaceutical analysis, they'll just basically spin up a thousand machines for eight hours, let all those thousand machines do work and bring it down. So no upfront investment. No upfront investment, yeah. And you find more and more people are trying to do, you know, big data has been, you know, a lot of vendors are throwing on big data and analytics and Hadoop and those kinds of things. And that's kind of where we see a lot of the enterprises going to use join for, is these things that... What is your take on Hadoop? So are you a Hadoop competitor? Are you a Hadoop partner? We actually can run Hadoop workloads on our infrastructure. And again, it's appealing because we actually can be very competitive with Amazon but we actually allow customers to, you know, you don't want your only Hadoop option to be buying a lot of gear or going to Amazon, right? So in our case, when you can put mixed workloads on something like Joanne that actually works with Hadoop workloads very well, it becomes something that's very interesting. And explain why you don't want to go to... You say there's workloads you don't want to go to... Well, I think a lot of it is a couple of things and I have a lot of respect for the Amazon engineers. I'm sure you do. But there's always a but, right? So in Amazon's case, I think it becomes a black box to people. They don't necessarily know the magic that's running the machine and Amazon doesn't let that stuff out. So they're clouds based on Zen and there's a lot of automation and replication type technology that they really kind of hold that intellectual property and they say just trust us with this and we'll run it. And some people are uncomfortable just trusting and I think something that kind of piled on unfortunately for all of us that really focus on the service provider space is that they had that outage. So the Amazon Web Services went down. Part of it was people said your application should have been built to be resilient, you should have had it in different availability zones. And the reality is like there were some bugs that happened that kind of didn't allow them to satisfy an SLA. You assume that you have your application split between two availability zones and everything should be fine well actually multiple availability zones went down. And that data was reliably replicated to different sites. That's right and they couldn't replicate everything and so it was just this cascading failure. The other thing is so people might not trust their information there and they want to know a lot more about the architecture. Amazon's still sitting in this weird space of infrastructure service but you assume there's some intelligence like platform as a service in there, right? You can just say okay give me a disk and make it available to me and make it redundant and then you just kind of say well I don't know what's going to go on behind the scenes but I trust that if something happens and it goes down that I'll be able to stay up. And I think people are kind of you know much about those things and I think that's why VMware has done so well in the off-prem and on-prem private cloud because people know it, it's still enterprise infrastructure that they've been deploying for years and they just feel comfortable with it. But it just doesn't satisfy everybody. The database administrators might not as be as comfortable because they want to know which spindle they're on. Right, so you know you just kind of find this interesting space of people moving and saying I want that plasticity but and VMware might not give me all of that that I want, help me find a platform that's going to work out. In our case we have outstanding IO performance because we use local disk and we use all of the available memory in a box for caching. We also have something called DeTrace and we've productized that in cloud analytics. DeTrace is a Solaris now Oracle technology, right? Now an Oracle technology. You know, open source. So using the open source of all that. Yeah, so to give you kind of a little backstory there years ago when Jason Hoffman and David Young started the company they didn't take money from anybody so it was very much you know stay as lean as you can, drive the cost down and so they I think they used FreeBSD many, many years ago and then probably about five years ago six years ago they switched gears so let's start using Solaris because of these neat things called containers that we can use. And as containers is just another hypervisor, right? Well it's really OS virtualization. It's not necessarily a hypervisor because we don't emulate physical hardware. So we take a lot of that workload away from that overhead and that workload away associated with that. The other thing is that when you're using a type 1 hypervisor effectively VMware is, you can't see what's going on inside that machine. So if an example is which is why you need all those tools like Veeam or stuff like that to see what's actually going on in the physical layer. Exactly. So in our case profiler and a bunch of those stuff Yeah, those guys, yeah. But we are actually able to see all throughout the stack from the hardware all the way through to the application stack in terms of latency. And so what you find is a lot of customers that's detrace. So customers who are conscious about performance or it's almost one of those things where I describe it as we call it kind of like an MRI. But the neat thing about the MRI is like if you're a football player and you're running down the field you get hit and then you go in for an MRI later, right? Well this is almost like I can constantly have that MRI. So when you make the impact so I can see as he's making that impact what's happening to his head in this case. They're actually building that into football hummus by the way. I know. Are they really? Diagnostics and sensors. They are, they are. But that was only an analogy. Go ahead. Yeah, absolutely. So we can do the same thing. So, you know in this case, your wildly viral Facebook game that you just created starts to have performance problems. You're able to actually use detrace and see latency and what database tables are slow. And it's a system spending time, you know, doing other calls that it doesn't need to be. And how can I optimize? Or is it just waiting for the network? Can you hear me now? Yeah, exactly. But that's huge, you know. You need that real-time diagnostics and information. So that's the, you know when you're using something like VMware, you actually it's a lot harder to get that because you started to layer something on top of that where you have to either hook into the hypervisor but with us we kind of give that to the customers all bundled in. The users can see it. The operators for the service providers can see it. So now you start to reduce the risk that you have when you start to go deploying the cloud. So you have your own, I'm sorry. No. So you have your own hosting environments. We do. And you're sort of a supplier to people who want to build their own. We are. And those people who want to create a public cloud. A public cloud or they can create, you know instances, private instances for their customers. So if you know, I don't know, like GM says, oh, I want a joint cloud but I want to have all these physical resources to myself. I don't actually want to expose anything to the internet because this is for something like analyzing GM sensor data on cars when they have car crashes or something like that. So EMC sold Mosey to VMware because they didn't want to compete with their customers. So how that's a very good point. So what's that mean for you? A couple of things. First is that we split the business off internally. So our service provider business is actually internally completely separate from software. It's in a separate container? It's in a separate container. Well done. That was nice. That was nice. Really good. I don't know what I'm going to be able to come up with next. We need to get points for that kind of thing. No, you should take away points. I need a marker or something. We'll play buzzword bingo. That's right. Well, after we get done talking about synergies, hopefully we'll get there, right? So we spun those things off internally. So we didn't have necessarily something relying on the other thing. We could actually focus on software development while at the same time saying this is our number one customer. And then things have kind of evolved. So there's a lot of companies that are interested and also maybe acquiring our service provider group as well. And we're completely fine with that. You're okay with that. Yeah. It's a source of revenue, but if somebody wants to go ahead and pay for it and use that as their cloud, they're more than welcome to. And the other thing that we do, too, is that it's very focused on the gaming and social media markets. Your own cloud. So that's why I mentioned Guilt Group and LinkedIn. Kabam, one of the biggest Facebook developers. The second, I think, behind Zynga uses our platform for all their Facebook games. So we try to stay verticalized. We don't necessarily go and market our public cloud because we don't want to compete with the channel. I lived there and I knew what it was like to well, being at EMC, it was always are you going to take this director? Are you going to let the channel go? And the answer was? It depends. If you close the deal, we won't. Yeah, exactly. I want deal control. So we lived, a lot of us had come from some of us came from telco backgrounds myself. I came from virtualization and storage and those things. So we're kind of used to working and partnering well with people. We make it understood up front that we're not trying to compete with people. If anything, it's kind of a nice thing because we already have seed customers. So if you are let's just say you're a company like Home Depot and you say I use joint public cloud for these workloads, but I'm not going to put everything there because I want, you know, multiple providers and stuff like that. Well, now we can go out to those service providers that are saying, hey, we're interested in spinning up a joint cloud and I can say, well great, because here's a customer. Wow, okay. That's interesting. So, you know, it's been an interesting ride. I mean just working for me was working for a larger company and I think the thing that sold me on joint was that their technology was truly different. So one of our biggest investors alongside Dell's a smaller investor, but Intel Capital came through and said those were the people. You sort of do listen if Intel Capital comes and says you should make a product out of this. We do a little bit. We like what you have and we're interested in helping you get that off the ground. So Intel's been a phenomenon. Yeah, and you know, some people may say well, Solaris containers, well, the first thing is that people who are deploying on our cloud don't care about the operating system running there. They say, I've got a UNIX-based workload or I've got Windows or Linux and I want to take advantage of these features and I really don't care. So go ahead and deploy. We can deploy actually, we can deploy Windows and Linux on top of our platform as well. I know, so that's in the open source community, right? It is. So there's interesting dynamics then that anybody else could go and do this then, right? Partner with Dell, take advantage of Solaris containers. They couldn't. In a lot of it, we've done things like put multi-tenancy in the IO stack, which has never been done. You contributed that into the open source community? I'm not sure if it's been put back yet, but the idea is since we're modifying that source code most of the stuff that we do is put back things around automation, orchestration all of our stuff has APIs. They sit outside of it. Yeah, so it's outside of that stream that's why you could technically pick up the software and go do it yourself but there's a lot more automation API accessibility and orchestration. Of course there's customer acquisition too. You've got a customer set there. That's right. So we can do that. We also have, so most customers don't care about the OS. Religion about OS is starting to slowly fade away. It's still there in the enterprise, but that hasn't actually happened at Apple yet. No, that's going to be pretty interesting. But you never know. So I shook you up. I'm sorry. I was just thinking like where's the, we have a Mac and a PC that's good. Yeah, we do. I just have to have to hitch my pants up to here. That's okay. I keep saying, well, you know Michael Dell's first computer was an Apple too. There you go. Did you know that? Is your next computer going to be a Dell? Well, of course. Of course, right? So also when the acquisition happened between with Sun and Oracle so there were some developers that developed ZFS, Dtrace and Containers that maybe weren't so happy and we're looking for other opportunities. So they're actually employed by you hired them. Yeah, so we contribute back with a couple of companies and we actually have the developers that help us make it happen. In MySQL, there's some sort of similar things happening around the MySQL too. Yeah, I think a lot of those things that went to Oracle kind of like, yeah, they get to have it, but a lot of the talent ended up leaving. Yeah, okay. It's unfortunate. It is. Things happen. So looking two years out where do you see yourself? Hopefully we'll see many, many giant clouds sitting out there running. You know, there's, as a startup obviously, you've developed disruptive technology and then other people are interested in you as well. So I don't know, we used to say what happens, but we're just kind of, we're we're just so engineering we're engineering focused and we just keep being down the path of making a pretty competitive product. Okay, so you talked a little bit about social media and you have your own blog. Tell us a little bit about the folks here, tell us the name of the blog and actually breathingdata.com. Yeah, so breathing data, I was just sitting there, I'm like, what do I do every day? And I just sit here and talk to people. I breathe data. I have to sit here and look at It's a nice phrase. It is. Is it relaxing or is it stressful? Depends. Depends on how much data, right? Or, yeah. When you start talking. If you're breathing data, you need a coach to help you breathe slowly. You do. It's almost like it could be a Zen like experience, right? It could be. So the things that I cover, I used to cover a lot. I've been really busy. I've only been with joint for about four months. So it's been a pretty quick ramp up for me. But I used to cover things around enterprise IT and the struggles that they have around. I focused a lot on VBOX when I was at EMC. And those were disruptive pieces of technology that the IT department didn't really like. So I would, I blog about things like governance and standards and, you know, how IT organizations can really start to meet the needs of the business. It's kind of funny because when we go in, a lot of people are interested in joint technology. And it's the business units that want to talk to you because they want to do things like I'm a major co who, I don't know. I have a really big website and I have a group that maintains that. They don't want to go through the same process as that internal IT has. They want to be able to develop on a platform they like. So I spent a lot of time blogging about those type of topics because when you go in and talk to somebody about VBOX the IT guys would be like, no, that's going to take my job away. I'm the one responsible for designing enterprise architecture. And the point that I always love to tell people is that you actually don't get paid to design infrastructure. That's just a necessary evil of your job. But it's the same thing of like building a PC. Would you rather build a PC or would you actually rather work on that PC? I'd rather build it. Sorry, you went to the wrong crowd. Well, I'm not a coder. But you didn't build your PC. No, I like to work on it and I like to build it. But you know, I have in the past. I'm not that specific one, of course. But I was the same way, right? Yeah, I like to build PC. I did and I kind of, what happened to me that made this transition mentally happen. You grew up. I grew up at some point. I know. I'll help you guys. We're here in the land of Disney World and you grew up. I'm here to coach you. Okay. This is all about reading. So I remember in I think 2003, I got an iPod for my birthday and I was a big Linux guy. So I had a Linux box and I had to try to get my iPod working with Linux and it actually took me like three days of like compiling kernel drivers and getting the right hardware. And it was fun to do that. But I'm like, I really would just like to use my iPod and listen to some music at this point. So I don't know. That's the kind of that's what I equate a lot of it to. There's certain things that you probably do want to build. Of course. And generally speaking, for sure, most people want to be able to just use it. Just use it. Or spend time building the things that are important to you. And not building those things that are supposed to be a commodity. I think I mean, you know, just take that high high level and that's what life is all about is focusing on the things that are important to you rather than you know, doing everything, getting involved with every little piece of your life. Like you don't have to hand crank your car anymore and that's probably pretty nice. You're not supposed to have to do that? No. But I did rent a car where you had to lock the door individually. Yeah, you did. It was an accident of a JetBlue bargain thing and I didn't need a big car and I wanted decent mileage so I got this car. You have to lock every door individually and you have to roll down the windows with the crank. It's really great. Congratulations. Yeah, my kids have been complaining about it. The whole trip is great. Well. No, it's fun. Yeah. So that's what breathing... Is it really fun? No. He's being totally sarcastic. So that's what breathing data is all about. How often are you blogging there? Not very often. Probably like once a month when I'm good. But you're going to do it more often now, right? Absolutely. I'll do it more often. Because obviously you can't have a blog and not be consistent. Right. Exactly. That's one of the things that makes it work. It's consistency. It is. I have a blog but I don't blog. That really doesn't come on very well. Well, thank you so much, Ed. We really appreciate your time. Thank you.