 So please help me welcome Jose Avila from CrowdStar. Hello, everyone. Thank you for having me here. I'm very excited to be here. I hope all of you are having a great time in Barcelona. So I'm the VP of Engineering for CrowdStar. And I'm going to show you a little bit about what we're doing. So CrowdStar, well, we were founded back in 2008 in the good old heydays of the Facebook gaming platforms. Yes, we're responsible for a lot of those requests you guys got. I'm sorry about that. We had published over 25 titles. And we had over 300 million users have been playing our game. In 2011, we decided to make a shift to mobile. And we created the first mobile title for women. It was called Top Girl. And it was a top-grossing game for quite a while. In 2013, we made the shift and decided that we needed to give more for the entertainment for women. And we created a covet fashion. So covet fashion is the first mobile entertainment platform for women that makes use of real brands and real items that you can purchase on any store around the world. It's a competitive platform, so what's that? Well, covet fashion allows women to dress up an avatar or a personality for events that we deliver to them on a daily basis. They create over 150 million looks a month. And then they submit those looks for voting on the platform. Over half a million women play this game around the world. And they vote 50 million times a day on these looks. What are you looking for? Well, everybody wants to have the top look in the game. And the competition is very fierce. Covet Fashion is published in iOS, Android, and Kindle. And it's available around the world in six languages. Now, what you guys really care about is how we get this running. So covet fashion has been deployed across four data centers in the US. We're using a multi-cloud infrastructure. And we have a combination of KBM and Vermetal and some of them. We use about 200 servers. And we process about 100 million API calls a day. Now, we've been developing this over three years. And a lot of challenges have come up. The first one was we were traditionally launched on virtualized instances, which are great for scaling up. But everyone can agree that it gets very expensive once you reach a certain scale. We needed to address that quickly. And we had to also, with the complexity of the stack, as we developed more features and things, our bootstrapping was starting to get very brittle and convoluted. Traditional bootstrapping takes a long time, has a lot of is-else cases and things like that. So it was taking a lot of time and making it very brittle. And since we operate with a very small team, we needed to be able to do more with less. And that's where Vermetal really came to change the game for us. Now, Vermetal, traditionally, everybody knows great performance, very affordable. But you have the issues of long-term contracts and long delays on provisioning all these servers. And that's where OpenStack Ironic really changed the game for us. With OpenStack Ironic, our guys have already been using a lot of the OpenStack APIs. They were able to change a few lines of code in our provisionings and be able to launch bare metal servers within minutes. By doing this, our first deployment, we were able to reduce our cost by about 60% and also reduce our latencies by about 40 milliseconds. So that was great, because our footprint actually got smaller. With OpenStack Ironic, the deployment was a lot faster than the traditional one day to a couple of weeks to get bare metal up and running. The next thing was our ops team was overwhelmed all the time by the developers asking for changes on the servers. Or can I have this library installed? Or oh, I want to use Node with this or this or that. And the bootstrapping, like I mentioned, was getting very complicated, very brittle. Sometimes somebody would change something on one of the servers. And it was just unsustainable for us. And that's where containers really came to really save us on a lot of these things. I mean, there's no silver bullet, but it's very close to getting there. With containers, we were able to give control back to our developers by providing a container to the developer in moving the application into this logic. The developers were able to start changing what they really needed, not depending a lot with the server that it was running on, making sure that libraries were there. They were able to modify the image. And then once they were done, our operations team would just have to check it out and make sure that they didn't do anything they weren't supposed to. Bundle it up and make it available on the repository. By doing this and using orchestration and a lot of other stuff, it allows us to actually create safer and simple deployments. With this, they were able to bundle up the images. Because they were running it locally, they knew that everything that they had coded and the dependencies of libraries were there. Once it was bundled, it gets deployed. And they are able to see that a specific version of the image is already running and that it has everything that they expected. And it's actually working the way we expected as well. Also, for Opsteam, our bootstrapping went down from a few thousand lines to a couple hundred. Just because now all we have to do is launch the bare metal servers, provision a few lines of libraries and code, and then it was up and running and available to run anything on our stack. The other thing is that it allowed them to start mixing different stacks of server that now could run a node implementation together with a lamp stack or a Ruby implementation. So it really created us. It gave us the advantage to reduce the number of servers we ran because before we had incompatibility on the OSs or the libraries. Now, all this wouldn't have been possible if we didn't have a group partner working with us. And that's where internet really helped us out. So internet is a very active member of the community. They were actually a finalist on the Super User Awards yesterday. And they really embraced OpenStack. They're contributing all the time. They have a lot of commits. And for us, they gave us access to over 15,000 bare metal servers that are being provisioned right now with a very broad configuration profiles. So this gave us access to raw power in the bare metal infrastructure to be able to run our things at a very affordable price. They have data centers across the US. And they have a great team that helped us all the time. So coming down for us in the future, we're actually launching our first competitive home design title. If I can explain that, basically, users around the world are going to be able to use real brands and real furniture that exist and you could go to a store made by some of the most exclusive designers out there. Install their rooms for a specific look that they're going for and then compete again the same way that they do with fashion. This is live already in Canada, and we're getting some great feedback. And it will be available mid-November around the world. It'll be published in iOS and Android. And we're very excited about it. It looks beautiful. Make sure you check it out once it's live. Also, if you want to know a little bit more about how we're doing this transition into bare metal using ironic containers and orchestration, we're giving a talk tomorrow at 5.30 on Room 211. I hope you guys will come and check it out. And hopefully, we can learn more from you guys. Thank you very much.