 Live from the Moscone Convention Center in San Francisco, California, it's The Cube at Oracle Open World 2014. Brought to you by headline sponsor, Q-Logic, with support from HGST, violin memory, and MarkLogic. And now here are your hosts, Dave Vellante and Stu Miniman. Welcome back to Moscone, everybody. This is The Cube. The Cube is our live mobile studio. We go out to the events. We extract the signal from the noise. This is our fifth year at Oracle Open World. Fifth year inside the Q-Logic booth, we had a sort of sneak in year one and maybe in five years from now, Oracle will give us a big stage like most of the guys do, but we've got a simulcast going on. We're here in the Q-Logic booth, also John Furrier and Jeff Fricker in the Cisco booth. So thank you very much to our sponsors for allowing us to do this. Stu and I are really pleased to have Jerome Nakaton. He is the CEO of Scality, entrepreneur, founder, serial entrepreneur. Jerome, welcome to The Cube. It's good to see you. Thank you, thank you for the invitation. So Scality, interesting company. We've obviously been on the share shift that we think is going to occur in the business. We're always looking for disruptive technology. Scality is one of those companies. You've had a knack for doing that in the past, but give us the sort of high level overview of Scality. Well, it's really simple. We think that the storage world is going to completely change and we're going to provide the next generation of storage. Essentially for the past 20 years, storage has been seen as a hardware play. You know, when EMC or NetApp sell storage, they sell big stack of racks and that's what they call storage. Obviously of the storage is that you can do it with standard servers provided that you have the right software on top of it. This is the way that Google has been doing it, the way that Facebook or Amazon are doing it today. It's much more reliable, delivers much higher performance, and it's much less costly. So it's all benefits. And that's what we do. We sell to enterprise and service provider. They want to control their S&T and continue to operate their own infrastructure. We sell them the software that put on top of servers will deliver a huge and efficient storage platform. I'm always interested in entrepreneurs the seed of the idea. When did you sort of have that epiphany that things were going to change? And then when and how did you act on it? So the seed is always something that is intractable and you know, it's in the air. For those who can pick it up. In this case, it was in the air from the people who are customers today. So in 2008, I went around the world and interviewed many telcos and asked them what's the next generation technology I can build for you. I know internet infrastructure. I know scaling technologies. What do you need? And it was really amazing whether it was in the US, in Canada, in Europe, in Australia, in Japan, everyone answered the same thing. We see the writing that eventually our competitors are going to be social networks and companies like Google, not other carriers. And if we're going to compete with them we need to be able to store consumer data in a way that's efficient, cheap, and reliable. And our suppliers are not giving us the tools to do this. Could you invent a new tool to do this? So really, I mean, it was as simple as listening to customers and developing the technology for them. Obviously, it's really easy to say it that it was simple six years afterwards. It's taken us six years to be where we are now. Well, when you start with the blank piece of paper, which is essentially what you did. I mean, you had obviously experience and you drew upon that experience. It does sound simple, but it's not. And we know firsthand, what kind of customers did you talk to? You said telcos, but how do you know you're talking to the right people? How do you know you're getting the right feedback? So we, in my previous company, Bizanga, we had been selling to the likes of Comcast, Time Warner Cable, Orange, United Internet, Telstra, and quite a few others in the world. And so I just went to see my customers. I had great relationship. I knew that they would tell me what they really think rather than what I want to hear. No sugarcoating. And that was great. I would say that when you're starting something new, in a way, not knowing anything about it is an advantage. So we didn't know anything about the storage industry. We didn't know anything about storage. I know a lot about internet infrastructure and scaling, but I didn't know anything specifically about storage. And it was the same for our CTO, George O'Reilly. So we looked at the problem really with the eyes of, what's the problem? Let's characterize the problem. And let's think about a solution from scratch. It's all white page of paper. And that's what allowed us, back in 2008, 2009, to design something that was truly innovative. And the funny thing is, I went to see a few of my relationships who are in the storage industry, and they said, what you want to build is impossible. It's never going to work. It's never going to deliver the performance that you need. It's going to have too much latency. I mean, basically, it's not going to work. And we just believed in our vision. And that's probably where, you know, an entrepreneur is different from everyone else. And we just said, no, we're bold. We're going to do it. And it's going to work. Yeah, well, that's interesting. You're back to 2008 timeframe. You had to have some foresight because, well, you could peek into what Amazon and Google were doing. It wasn't clear that that was going to seep over to the enterprise. So you had to make that bet. And you talked to customers. That's always a good idea. So, okay, so you funded the company. Maybe talk about that part of it and sort of take us through where you are today. So there's one originality with Skilletease. We started this company really on two sides of the notion. We have the headquarters here in San Francisco. This is where we manage the strategy of the company. This is where we manage our relationship with the big partners, all the IT industries here, pretty much. But we have the engineering based in Paris in France. And we've done that from day one. At the beginning, there were just a few of us here in San Francisco office. And we're growing the engineering team in France. And the reason we've done this is really to catch the best of both worlds. Definitely Silicon Valley is the center of tech innovation. Nothing can give you as many ideas and a better feel for what's happening in the future than being here. At the same time, it's really difficult when you're here to get the best engineers. I mean, they have a tendency to go to Google and VMware and Facebook. And if you're a new startup, and if you don't have a track record within Silicon Valley and a prime AVC with you, you just don't attract the best people. In Paris, you've got amazing engineers who are about half the cost of Silicon Valley engineers. We currently pay Paris engineer about half the cost. And actually, a better productivity. So it was an easy choice. So we've done like this. And we initially financed the company back in Europe just because it was really easy to get money. There's a lot of help, especially in France, but also in the UK, for new entrepreneur and especially high-tech entrepreneur that have a large vision, which we did. OK, so where are you at today? Maybe give us some basic metrics, a head count, maybe talk about some customers, maybe some proof points. So we've got 50 customers around the world. We're very active in the US, Western Europe, and Japan. Some of these customers are flagship names like Comcast, I'm on the cable, Orange in Europe, Tillinette, and a few in Japan that I cannot name. But about a couple of years ago, we decided to go beyond our initial market. Initially, we were selling to Telco. So now we're in the enterprise and government market. We want to deal with the Los Alamos National Lab here in the US, which is a major deal. They're envisioning deploying 500 petabytes of data on our system. I mean, this is huge. I just have to interrupt. Oracle has 400 petabytes in its cloud. You have one customer with 500 petabytes. That's a big number. It's definitely huge. And also, we have a lot of customers that do a video distribution using our storage. Customers like Deluxe Digital here in the US. And Daily Motion, the competitor to YouTube, largest competitor to YouTube, actually, does its storage now on Scality. So that's also a great win. In terms of the idea of a company size, we are 90 people, 30 of them in the US, mostly management of the company, and sales and customer service for US customers, and then 90 in Europe with the development and sales and customer service for European customers. And we're just opening very soon an office in Tokyo. Great. So Jerome, if we look at the storage market today, we see a bit of a splitting. At shows like Oracle Open World, it's all about performance. So all of the all-flash array guys are here talking about in-memory databases. There's certain applications where performance and latency are key. But then there's the capacity side of the market. And that's huge growth. I mean, we've talked for years and years about how unstructured data is outstripping all other growth. We're seeing the Skelloc NAS companies are doing quite well. You guys play really in that capacity play. And scalability that only a couple of companies can do. I mean, when you're talking about companies like EMC, talk about their petabyte accounts and things like that. Can you talk a bit about what does scalability mean to your customers and how you approach that? So we have only 50 customers, but almost all of them are petabyte accounts. And that's our approach to it. So yeah, historically in the storage market, basically you had a product for each usage profile. So you had a product for database high performance. You had a product for video, a lot of read. You had products for a lot of write and read never. And you had five or six products in a data center to be able to do the complete storage for a given enterprise. Our vision is that in the future, this is going to radically simplify. And you will need essentially two products. You'll need something that delivers extremely fast IO. And this is what you're seeing here on the show at Oracle Open World. And this will be all silicon based. Today it's all flash. There may be other generations of silicon, but it's going to be all in memory in some form of memory. And that will deliver for the workloads, typically very heavy relational database, high frequency trading and specific application that really required these IOPS. And that's probably, you know, I'd say about 20% of the whole data stored, maybe 40% of the value of the storage industry. And then there's the rest. The rest is the other 80%. All the videos, the music, all the thousands and millions and billions of documents that are generated everywhere. All the backups and archive that you need to have maybe for compliance. The rest of us, all the rest of us will go on capacity system like the Skeletee one. And one of the uniqueness of Skeletee is that we can address a very large panel of workloads within the same system. And that delivers huge cost reduction by having one system that takes 80% of your workload. All right, so Jerome, we've been following you guys for quite a while. When you started the company, there was no software defined storage. Flash was quite early in coming back to Enterprise IT. Today, you guys are positioned as a software defined storage company and your software, right? You don't have hardware that you guys make. So, yeah, a couple of years ago, I actually did an article about everything's cloud, don't believe it. I could probably do an article on everything software defined storage, don't believe it. But it happens that we are truly software. We don't sell any hardware. Well, yeah, but just because you have software, doesn't mean that you're software defined storage, right? That's absolutely correct. Yes, there are many hardware companies that also, you know, claim. So, the point here is two things. First of all, we're totally hardware agnostic. Our customers buy software from us and they buy whatever hardware they want from whatever manufacturers. We work very heavily with HP, Dell, Supermicro, and Cisco, UCS quite a lot as well, and really with any hardware manufacturer that is presented to us. The other thing in software defined storage is that the whole system can be managed by API. The holy grail is that tomorrow it's going to be application that will self provision the storage in terms of amount of storage and quality of storage that they need. And that's what we're delivering. So both a system that's completely software based and a system that can be driven by application and respond to application requests. So, I'm hoping you can unpack this a little bit for us because if I look at some of the software players in the software defined storage market, just to build kind of one of those to get the integrator there, to get the hardware to test it with an application and to build that up seems to be slowing down some of the growth in those deployments. When I talk about doing that at scale, isn't that harder or are you focusing on specific applications and specific use cases in which case you have a template and you can go out and do it? So, I mean, definitely in customer conversation we sometimes have the worry about integration. But the truth is we don't require any integration. I mean, deploying our software on any server is like pretty much deploying Windows on a laptop. You don't really think in terms of integration, it works. So, this is a conversation that we sometimes have early in the sales cycle but it really, truly disappears and no, it's not an inhibitor to sales in any way. The one thing that can be an inhibitor to sale in which we have solved that security is that you need to present to the application the interfaces that they expect. So, some of our competitors that also have a software product basically think that tomorrow is already here today and we don't think so. To this application expect to see a file system for most of them. I mean, according to IDC, 84% of the storage that is deployed today expects the file system. 84%, this is huge, okay? So, we have been able to deliver to the market a software-defined storage platform that actually presents NFS or CIFS and can also present an object store if that's what the customer wants with compatibility between the file system and the object store. And so, the proposal that we have to our customers is deployability today, you get a system that talks to your application the way they are today that has the architecture of the future that is completely scalable and reliable and a low cost and traditional architecture. And at the same time, we give you a migration path because you'll be able to change your application over time and we use the data in place without needing to do any data copy or data transfer. Yeah, Jerome, great point because I think you specifically talked about object storage because when Scalely first came out we thought of you as an object company and it was what I used to call that air gap between the way everything's been written today and where we all know the future is going. I mean, if you look at clouds and how they are being built, object store and data stores is what they're going to have. So, your telco clients, are they moving over to object faster? How are we on that transition of workloads and applications? Let me clarify, we're still using the same technology as when we came out in 2010. So, the core technology to Scalely is an object-based storage technology and we inherit all the nice property of object-based storage. But indeed, we see ourselves as a general purpose storage for the 80% of the workload that's not going to go to flash or in memory database. And so, our customer base, 75% of our customer base use our file interface rather than our object interface. So, when does that change? Or does it ever change? So, I think that file interface is there to stay for a long, long time. So, I still meet customers that have programs that work in COBOL and they're still in production. So, I'd say the horizon is at least 20 years. What's going to change much more quickly is the way people think about storage and the way they deliver storage. I think that software-based storage, the kind of technology that we're proposing to the market, is going to be a big shift, probably two to three years from now. What we're doing is a little bit like, bringing TCP-IP to the market in the mid-90s where the good networks were either Token Ring or Novel Network. And when TCP came around, was like, yeah, that's the future, but the future is for a long time from now. And that's the point at which everything switched. And it was over a matter of about 18 months that Token Ring became totally out of fashion and everything became TCP-IP. Right, it's like trying to time the stock market. It's impossible and then it just happens very quickly. So, and you know what we think about this share shift, we forecast it out and think it's going to be pretty dramatic. Timing it is somewhat difficult. So, but now you started this segment off talking about that your true software company, storage is, you know, you have a different angle on storage, traditional storage companies sell you a big giant iron box. Now, having said that, all the big storage guys will say, well, we have all software engineers. We're all software guys. Now, I want to pick your brain, love to get the perspective of an entrepreneur who's trying to change the world. How do you see that affecting the big guys? I mean, you've seen EMC come out with its Viper and the software-defined storage strategy and everybody's talking, software-defined storage washing. How do you see the big guys responding? Can they respond effectively? Well, I think it's the famous innovators dilemma. When you're a big guy, you have a revenue stream you need to protect and you've got a big chunk of your organization who has a vested interest in protecting your historical product. And then you set up a new team that's going to develop the product of the future. And, you know, as a CEO, you know that this is where you need to go, but you've got the rest of your organization that's essentially fighting against it. So I have no doubt that the established players can get great software engineers. Even if they do get the great software engineers, it takes several years to develop the kind of technology we've developed. So, you know, they started maybe a year ago, it will take them three, four years to be able to do something similar and we're not staying still. Okay, we're continuing innovating and we have quite a lot of nice research that's not out on the market yet, but that we have in our research lab. So I think that the big guys are going to try to do it. I think they're going to establish teams to do it, but I think it's extremely difficult for them to innovate at the speed of a startup. Do you think we'll see another storage company emerge that is able to stay independent and stay viable and stay big? I mean, let's see, was NetApp the last one? Yeah, NetApp is the last one. I've looked at this over and over again. Yes, NetApp is the last one. It's about 20, 25 years ago. So I would say this is our ambition and strictly speaking, statistically, the odds are against us. And I think, I mean, we definitely have a big enough market to do an independent sustainable company. So the market is carrying us to be that. The question is, will the big guys let us become independent and big? Well, so what were the characteristics that allowed NetApp to do at big market? Innovative technology, good funding, obviously, and ability to execute? That's actually all easy. You need to be carried by the market at some point. Which they were. And, well, they had the web market and the virtualization market, so they had two markets that were really carrying them. Tailwinds for them, yeah. And we think we are going to have that. I mean, we see it coming with software defined storage. I mean, the video customers, we didn't have one customer doing video on our storage platform a year ago. We have six today, and this is growing super fast. So we're seeing the tailwind come. But we need to get really big, really fast before others may snap us out. Scale, all right, you got to scale fast. Okay, that's good. Jerome, can we talk about your move to the enterprise? Because if I look at the service providers and the telcos, they're storage team and the way that they think about infrastructure is different than the enterprise. Maybe you have a different experience. What I've seen is service providers, first of all, their growth pattern is they're willing to try new things a little bit better, and they're not necessarily as siloed. The enterprise has typically a storage team if they're big enough. What are you seeing? Obviously your customers are ones that are trying something a little bit different. So how do you scale into the enterprise because is the storage team an inhibitor to your growth? So we initially went to the service provider market because they had a bigger pain than the enterprise. I mean, they were already at the point that they were storing petabytes on application that just couldn't bear the cost. When we go and meet with an enterprise like a bank, for example, cost is actually not a big problem. So they're telling us, okay, we have these big things. Yes, it's bigger and yes, it's complex, cumbersome. And yes, we pay a lot of money, but at the end of the day, it works. And so to really capture the attention of the enterprise, it's not enough to say, we're going to reduce your invoice by half. You know, they say, okay, it's nice, but what am I losing? Where the enterprise is really interested is when you simplify their business process and help them move their own business process towards the future. And that's what we're really talking about. When you're able to essentially destroy silo data so that one application can talk about the data of another application in a way that is easy. When we can decrease the amount of operational steps that are necessary to upgrade a platform or to modify a platform or to add a new application to the platform. So essentially it's about making their business faster. And that's really the conversation we're having with the enterprise. Today, I would say our shift to enterprise is remarkably easy. We have conversations in many verticals where they welcome us, but they welcome us on a business discussion. All right, so what are your biggest inhibitors to growth then? Right now it's recruitment. So anyone interested to come and work for a multicultural, multi-geography startup, especially engineers, please come and apply because we're really looking for more of you. I could go on here, we're getting the hook. I mean, the whole dynamic is really interesting. Actually, I got one last question for you. So we're here at Oracle OpenWorld, the heart of sort of traditional applications and IT. 10 years from now, what's this business going to look like? The IT business isn't going to look like the cloud service guys today, the Amazons, the Googles, those guys, or is there still going to be a big flavor of traditional IT that starts to adopt those kind of techniques? What do you think? So I think that, if you don't mind, I'm going to respond 15 years from now. Great. Humans are slow to change. Okay. 15 years from now, I think that all SMBs who are consuming a large share of the IT business, they will be using cloud services, not so much in the form of an Amazon infrastructure service, but rather like Salesforce. So think about many applications. Think about all the apps that you have on your iPhone. That's going to be IT for the SMBs tomorrow. It's going to look like this. And then there's enterprise IT, you know, the Fortune 500 company or the global 2000 companies. They will still do traditional IT in the sense that they will want to control their applications. They will still be, you know, Oracle Database and big applications, but they will be forced by their own user to present it more in an app style. So they will have to produce this IT more in a cloud style. So think about enterprise IT 15 years from now, more like Facebook. And that's our vision. Jerome, thanks very much for coming on theCUBE. Congratulations on all the success. Great entrepreneur, great leader, really clear thinker that appreciate you coming on and sharing your insights. So good luck. Thank you Dave. Thank you Stu. All right, keep it right there. Everybody will be back right after this. Live from San Francisco at the Moscone Center in the QLogic booth, stop by and see us. This is theCUBE, right back.