 Hey everyone. I'm Yorm Heller. I'm a COO and co-founder of more Flabs. I prefer to make this interactive so feel free to heckle me if you want. You can just yell or jeer or throw your phone at me if you want. I was going to have Clinton throw these hats at people, but if you want to grab them just feel free. There you go. So today I'm going to talk, well first a little bit about who more Flabs is, but we're going to talk today about how you build cloud infrastructure specifically for Asia, what our experience has been, and ultimately how you compete with AWS. How do we take open stack and actually enable competitors with AWS, which I think is the opportunity that we're really looking at. So who's more Flabs? More Flabs delivers fully integrated software products that deliver infrastructure as a service. So the way to think about us is we're ultimately a software vendor, but working with preferred hardware vendors like NEC, we optimize their hardware so you have a solution that enables you to compete with Amazon on price and performance. We're headquartered in Los Angeles, but what makes this also really different out here is we're actually Asia-centric and Asia-focused. So we have an office in Tokyo, two offices in the Philippines and an office in Singapore. We're pretty well venture funded. We're a founding gold member of the OpenStack Foundation, so we've been here for a while and plan to be here a lot longer. And we've been around open source for the better part of 15 years, so we know how to work with open source community and actually build products that are meant to be consumed by enterprise and service providers. So just kind of quick highlight, what we try to do is drive price and performance, right? So we do this by taking a systems approach, which again is pretty different than most of the other vendors out here. We actually really focus on optimizing that hardware, and we also have a modular framework so you have building blocks. So we're very opinionated with how we do things and we're going to tell you how you want to build your compute and how you want to build your storage and we don't want to worry about all the other options. We productize open source for service providers and enterprises. We actually integrate billing into our platform, so service providers who come work for us, this is really a soup to nuts platform. It's all margin optimized, so when you take in the hardware costs along with the software and data center costs at an Amazon price point of around $43, you're looking at 50% margin. We can tell you how we do that at our booth, and really we're looking at this as the first step. So we're trying to have the entire software defined data center starting with infrastructure as a service for private and public clouds. So I don't know how many of you guys have seen the wire, any hands here? No? Very, very, very few, right? Okay. So there's a scene here where this guy gets out of jail and he's been in jail for a while and he comes back to the streets and he says the game done change, right? What he's trying to say is things have changed and he doesn't know how to work in this world anymore. And what we're experiencing infrastructure today, specifically with Amazon here and with OpenStack, we're disrupting the incumbent players. Things are changing, right? And the leaders of yesterday are not the leaders of today. And there's a lot of opportunities for us, specifically in Asia, to carve out our own niche so we can become the leaders of tomorrow. I'm sure some of you guys have seen this, and a lot of people talk about Amazon Web Services here, but they are massive, right? This cannot be underestimated. This is affecting everyone. So if you have a service provider who's in Japan or if you have, you know, your own enterprise who's been working to date with, you know, a specified hardware vendor, they're all moving to Amazon, right? They are dominating this cloud infrastructure market. It's more than five times the capacity of the next 14 largest providers combined. That's crazy. And that's only starting to accelerate. So you're just starting to hit that curve right now. And what should be really scary to everybody here is they're starting to partner. So they're taking their own infrastructure and I'm sure a lot of you guys have seen this. They're actually applying this to things like the CIA cloud. And ultimately the faster AWS grows, the less room there is for everyone else here. But there is still opportunity, right? So this is a graph with the APAC service provider cloud market. So if you look at the blue area, you see just general infrastructure as a service. That's grown pretty well, right? But if you look from 2013 to 2016, you see that red area growing tremendously. That's hosted private cloud. This is a fundamental difference with the Asia cloud market. So you have a lot of enterprises who instead of building their own private cloud are going to service providers so they can get a hosted private cloud. The reason for this is there's a skills gap. There's not enough people here who know how to build private clouds themselves. So they have to centralize the expertise in one place. So they're fine outsourcing their infrastructure. And to further this point, there's still a growing demand. So the way you should look at this is this question is how likely is your organization to outsource your infrastructure? So this is for applications and infrastructure generally, right? If you look at the blue area, that means it's a priority. In Indonesia, I mean 66% of the enterprises want to outsource their infrastructure. Again, think about that. That's a pretty crazy market. This is underserved. They're looking to actually outsource their infrastructure. They want to do this, but they're still not options, right? So the question then is this region is tool in right now. The ASEAN region is tooling. So how do you become the enabler, right? So you know that they have a skills gap. You know that there's an opportunity and how do you take open stack and actually deliver on the promise? So let's go back to Amazon Web Services, okay? This is their disruption recipe. Everybody kind of knows this right now. You start with really cheap hardware, right? You might work with an ODM like Quanta or somebody like that. Open source virtualization. I think they started with Zen before. It's probably some modified version. And then their big innovation was operational automation and then put an API in front of it. Super easy, right? So what do we do? So we start with the same idea. So you want to use optimized hardware. So modular implementation. So you have a building block approach. Things have to be homogenous. The more that you have different things, it doesn't scale. The more simple it is, it scales. If you have complexity, you can't have the operational efficiency to work at scale. And you have to start at this level, even if you're a software vendor, because you have to compete on price and performance. You have to think about this from the beginning. If you start with the software and then think about the hardware later, you'll lose. It does not work. Then have a really opinionated approach. So there's a lot of people here who are doing open stack and there's a lot of different ways to do it. But ultimately, I'm here to tell you that it's more or less the same thing. You're going to get open stack, right? And you're going to have virtualized infrastructure. And some people will say, I use Chef, I use Puppet, or I use Fuel, or I use JuJu. Who cares? Just have one way, figure out what it is, optimize that approach, put it together, and have an approach where you can do that repeatedly over and over. And what we've done, and what I think is actually really important is you have to improve the usability over AWS. So Horizon's a really good starting point. But for us, it's a reference implementation of open stack. So you actually, we built our own UI using Ruby. We can show it to you at our booth. I don't want to bore people going into this right now. But we think this is a better way to do it. And we think about the user instead of just thinking about every way we can leverage the APIs to get this done. So when I talk about an opinionated approach, this is kind of what we do. So we start with NAC hardware or Dell hardware utilizing Intel SSDs. We have some magic that we do in between the actual virtualization layer. So we actually optimize the hardware and optimize the hypervisor. We use the Ubuntu distribution ourselves. We've been using Accenta today, but we're really big fans of Ceph in a company called Acala. And then we have our dashboard, which has built in billing and VM monitoring. Again, just really important, take an approach, have everything integrated, and you want to just have it all work together. If you start trying to piecemeal, put it together, it takes a really long time. Our CTO, Christopher Ado, who's sitting in the second row, can tell you all about that. It takes years. It sucks to start and do this. So let us help you. So what makes this unique, right? We call this hyperdense. So start with your optimized hardware architecture. Figure out how you can stick as many virtual machines in there as possible while preserving your performance. You want performance to be really high because that's how you differentiate. So cheap VMs that are actually faster. And then you want to do best of breed software. We call that mCloud. So a complete systems approach, even as a software vendor. So you want to think about delivering these virtual machine economics. How much does it cost to deliver a virtual machine per month? You have to think like that. These are some pretty interesting metrics, right? So we put our money where our mouth is. So we ran some benchmarks. Unix Bench, pretty standard benchmark that you'll see. Amazon's pretty low, right? 207. Running on NEC hardware. This is about eight and a half times faster. And we have, what, 12 times the IOPS. At that suggested price point of $50 per vCPU, that's well over 50% margin. I think that's pretty compelling. We can talk about all different kinds of open stack features. This is what I care about. And I think this is what we should all care about. So this is something that we actually have on our website. This is a configurator. So it just allows you to look at how many vCPU you want and then how many terabytes you want and just configure your cloud. You shouldn't be getting into the weeds with how to figure out a deploy open stack and which kind of hardware components you want. So the cloud service providers that we've worked with to date, they're not the same as the US. They do not have the same skills, right? So you have to have this modular approach because they need help to figure this out. So starting with a modular and scalable solution that's pretty small and then something that addresses private cloud as well as public cloud. So in our architecture, we have the same building blocks for public and private cloud. So a compute node, a controller, storage nodes, all the same thing. So we're able to start really small with something we call a project server. This is 1U. You can hear, again, more about it at our booth. So you have a 1U implementation of open stack. That's it. Super easy to get started, both for enterprise and service providers. Make it dead simple. We don't even worry about just having the software. We want to have the complete experience. And from there, you want to be able to scale to a really big public cloud without having to re-architect your solution with built-in billing or a private cloud. Same building blocks. You have a controller, compute node, and a storage node. So how do we do this? How have we been doing this today? So we've been starting with large US-based service providers like MediaTemple, which just was acquired by GoDaddy. I think there are about 50,000 VMs. So we're using this experience and actually exporting it and regionalizing it here. Once you have this experience, you're a lot more credible when you go here and you know what you're running into in terms of issues with scale. So working with large telcos, large service providers who actually already have entrenched enterprise customers and allowing them to deliver cloud services to them. So just kind of quick summary. And then anybody can ask questions or visit us at our booth. So we have a strong focus on APAC service providers with hosted private cloud. Again, there's not as many developers out here. And those are probably already going to Amazon. So differentiate by focusing on the enterprises. AWS cannot be everywhere. And we know that the NSA may think that they're everywhere and listening to everything. But that's actually helpful for other service providers who are here. You can differentiate because you can have your own support and you can actually have your data inside your country. I don't think that used to be as important. I think that used to be a little bit of a misleading sales marketing thing. Actually, now it's pretty important for people. So service providers have that strong strategic advantage. Focus on that. Enable service providers in Asia to compete on two metrics, price and performance. And how do you do this? Eliminate the skills gap. So make it super simple. So modular solutions. You really opinionated with how you deploy your software. Don't focus on all of the other different deployment options. Ultimately you're going to get to the same place. And then really focus on providing a back end for one country. So you can concentrate that in one service provider as opposed to selling into a lot of different telcos and a lot of different enterprises. And then it goes beyond that. You have to train for the support services, the training services, the marketing. It sounds like a lot because it is a lot. You can't just stop at the software. You actually have to train for the entire stack. So with that, thank you very much. Feel free to ask any questions right there. A little bit louder, sorry. Oh, that's actually a great question. So that's something that we've been... So I'm going to kind of repeat it back and you tell me if I heard it right because I can barely hear you. So do we focus specifically on applications for these markets? So we can talk to you all about running SAP, running Pronto. I actually wanted to include a slide about that, but I didn't think I'd have enough time. So that's a really important piece. So if you can focus on, let's say, the ten most used apps in these regions and enable an ecosystem approach. So ultimately you're still going to focus on delivering the infrastructure. But if you can provide the... Or find the preferred SAP vendors or the preferred, you know, Oracle vendors, whatever it is, and bring them as an ecosystem, that's really important. So if you just focus on the infrastructure, it's not really enough either. That answer your question? Great. Thank you guys very much. I don't know what our booth number is, but it's over there. Straight ahead.