 Hello, I work for Airbus, talk 2 seconds about what is Airbus, maybe some of you know us, yes, already tried, bye, nobody, so I'm an IT tool start-up leader, you will see why we talk about start-up, it's very strange, a billion dollar company with a thousand hundred of employees with start-up, very strange, you will see why, because we talk about debuts. So in two slides, because we have not bought our product, we have 50,000 employees for manufacturing plane, so it's only manufacturing plane, but we are a group that also build stuff for space, we build spacecraft, we build launcher, military and also helicopters. Ok, our revenue 40 billion, it's only for the civil aircraft, and this figure is the most important, 9 year backlog, so we have enough to deliver during 9 years, so one of the issue is when you buy a plane today, we deliver it to you in 9 years, it's incredible what is the solution to deliver quicker, and if you need to deliver quicker, you need to get better tools in terms of manufacturing and also IT. And we talk about DevOps, which is for delivering quicker, and we talk also about solution like OpenShift, which allure to deliver quicker. So this is the main stuff, 9 year backlog, I will give you another example, it's a time when you want to introduce a new type of plane and when it's available. In 2014, we have announced to the public that we will have a new plane, the 313 NEO, ok? Two years after, this October, the first flight will happen, two years between the announced to the public to the first flight, ok? We start the test on this October and we deliver the first plane one year after to a company, so approximately in 3-4 years, you design the plane, you do the test and you do the first delivery, ok? Don't be aware, the plane is secure, it's not because we go fast, don't be aware, ok? Ok, so I work for IT for the company and I mean, because we are in the US, we are also in the US. Since this year, we have a manufacturing in the US, we manufacture plane in the US, ok? It's in mobile, we have a manufacturing, we deliver few planes from this manufacturing since this year. So it's really important, we are not a European company, sorry, European and also English, we are also, we are more and more a world company. We manufacture in the US, engineering in the US and also in China. This is really important, so this is where you will find all our IT for Airbus. We are only approximately 1000, so it's not a lot for the IT store. And here, I will talk now about my domain, the infrastructure. So I work in the infrastructure, I support all the infrastructure for all Airbus worldwide. And this is the sizing of our infrastructure. So it's quite big compared to the number of mailboxes, we have a mailbox approximately 2.5 GB per user, this is the quota. When you multiply it's a terabyte. I would like to highlight just full point, it's quite fun but we have a project to replace all the mobile phones. When we replace mobile phones, we do roll out for 45,000 mobile phones. So the supplier that sells the mobile phones is quite happy. But when we do something to the size of the company is big. So when we do open source, it's quite big, it's impact all the company, it's a major decision. And I would like to go to the back office stuff, why? To the last figure, we do high performance computing and if you look the top 500, our computers are inside the top 500 for the supercomputer. We were well placed few years ago, unfortunately the battle is very strong and we lost our place. But we are still in, just for the story, servers, we have servers, a lot, but we have also servers in plain. If you took the 380, you have approximately half a ton of server in the 380, half a ton of server in the plain. So it's approximately 6 passengers, 6-8 passengers. They prefer the company to replace the server to passenger, it's more interesting in terms of economics. So I will move to master. So today Airbus have decided since approximately one year to move to open source. Why we move to open source? We get project, it was a 90SM project, and we decide to use open source to go. Mainly today, you know, there is no licensing, there is nobody from Microsoft, Oracle and so on, it's okay. We don't have to pay licensing fee, we have only the support to pay. Okay, so this is clear that you can do economy, it's in the beginning. After us, what we target is not that, the main point, is the top. Why the top? Because today, we think that the cost of the project will be not the technical stuff, will not buy licensing, it's human. It's a salary to pay to the human. So more you get people productive, more your cost will reduce on the long term and so on. But for that, we have decided to create a kind of inspiration, methodology internal from the open source. We call it the open way of working. And this is linked to DevOps. So we don't go straight to the end to say we do DevOps. We say we change our way of working to go to DevOps and to push the technical stuff. Because the problem is what I say all the time to my Red Hat expert. They already sell us technology, but we don't use it. And we use it in five years, the time that people adopt the technology. So why sell us products that we never use, we use in five years. So what we focus on with the open source is a collaboration, transparency and sharing empowerment. For us, the most important is really the collaboration. With the collaboration and to break the silos. All of us, we talk about silos is to break the silos. DevOps is break the silos between the dev and the ops, how you can do it. And this is today our main point, our main work. So we are starting this work in terms of changement. But in parallel, we put in place OpenShift to support it from the technical point. Okay, next slide. So I will give you two examples of projects that we have deployed from this methodology. And that for us has a major success. So itop is one example. It's a tool that we use to support our ITSM, our ITIL process. This tool supports all the assets for the company. And we have deployed it in three months. In three months, we deploy a tool to support all the assets for the company. Server, back office, front office, connectivity, all the figures you have seen. Only in three months, in three months, we have deployed the tool and we have done all the change. That means training for a user and so on. Why we have successfully done it in three months? Because after one month, because we use an open source project available, we create a partnership with the community. And one month after, the first release was available for tests for our users. And we done one release per week. This is not possible if you don't do DevOps. This is not possible. So this is the main reason we have used it. For this one, for example, after we move to OpenShift. Another example, which is just for us, is a major success. Because the entry of services was in June. This is a portal for the end user. When we do a portal for end user, for 100,000 users. Okay? It's a portal where the end user will find all his IT support tools. How to manage your incident, how to manage your request. When we talk at incident at Airbus, it's approximately 80,000 incidents. Par month. After, you go to million per year and so on. And this tool is used by the end user to manage their incident, their request, their knowledge, database access and so on. We have deployed it, you can see. In two months, we get the first release available for the end user. Every week for 5000 approximately end user. Okay? Missing one zero. We open it after three months. After, we've moved to 4000. And in one week, we have opened to 4000 to 100,000. How we've done it, we've developed. Without, it will be not possible. Okay? And again here, we have used LifeRay, which is an open source software. We have relied on LifeRay. We talk with the LifeRay community and so on. So all the time, for all our projects, we really have strong collaboration with the community. This is a critical point for us. Our DevOps, no, I will move really on how we set that in Airbus. It is quite simple. We adapt it to our process. If it's not adapt to the process, nobody will use it. If I talk about OpenShift, if you buy the OpenShift, you put it somewhere in your IT environment. And you say to the user, you can use it, nobody will use it. If it's included in your project methodology, they will use it. So we have changed, we have adapt, change, work on the project methodology, on the tool. And we have included all of that going to OpenShift in the tool. Why? I will explain the first part. We use ITOP as an internal project referential, because we need some few information that you don't find in other software for managing projects. And from this tool, they say, I want a project. And when they click, it will create a repository for their project in Tulip, which is an open-source software to manage projects. I don't know if some of you know this tool, Tulip. It's well used also in US. And they create also their Git repository at the same time, and their project in OpenShift. So that means, when the project manager starts, with one click, he can create his project repository, the repository for Git, and also for the OpenShift. And it's fully linked to the process. Okay? This is okay for us. Nobody will adopt the solution. They have to raise requests somewhere and so on. The process must be flexible and adapt. After, I would like one word about OpenSource, which is really fun for me. It's two examples. Tulip, for example, we use it here in Airbus, and the community of Tulip and the company linked to the project don't know it. It's used by a lot of engineers, and they don't know that we are maybe one of the biggest users of their software. This is incredible, and this is OpenSource, unfortunately. When I work around the partner, and I ask to some of the project, do you know which are your users and so on, they say no, we don't know. We just know that we are downloaded hundreds of thousands of times. And which is quite interesting also, it's a GFrog Artifactory, I don't know if some of you use it. I've contacted the company GFrog, and I discovered that they have a representative at one kilometer from my building. So I contact them at the US, and I say we can meet, go to my building, and they say no, you can come, we are at one kilometer from your building. This is OpenSource, sometimes you never know. So for us, the solution, as I say, is an open shift with the pass. We use for two scenarios. The first one is a new project to migrate, to have them on an open shift. And the second is a migration of application, of PHP application we have. We have all the server and so on, and we need to migrate all this PHP application to something more flexible for the future. And we find that OpenShift is the best solution. So, unfortunately for Red Hat, we are not the major customer for OpenShift. We hope that we will become. But it's really a use case for the future and obsolescence. At the beginning when we talk with Red Hat, we are starting with obsolescence. And now we've seen that there is a huge opportunity and we go for the future. For us, today OpenShift and the container are the solution. We look into a new cloud solution, bigger, where we have the privates, the public, and the OpenShift. So, it will be included in our cloud strategy for the future. So, it's my last slide, my conclusion. With Open Source and DevOps, it is clear that we deliver quicker. I will give you one example. When I talk about the portal for the 100,000 users, we do the entry of services on Monday on the Friday I was with the project manager. And the project manager, c'est to me, is the first time that I will go on weekend and have nothing to do for an entry of services on the Monday. Because usually, when you have an entry of services on the Monday, you spend your weekend to work. And it was the first time that the project manager is something that is a 50-year experience. C'est to me, I will live and I will have nothing to do. So, this is clear that we can improve, we deliver quicker, with the higher satisfaction, with the continuous relief, deployment and so on. We reduce our cost because we have less bug. We don't discover the bug at the last moment. This one is particular, it's for Red Hat, thanks, because we need the right partner for making this project. There are two kinds of partners. The partner for the technology, selling the technology and selling some high consulting, like Red Hat. And there is a partner that we work with everyday, like Altramp for example. So, you have two kinds of partners. You need to find the right partner. The technology is not enough. You can have the technology, if you don't have the right partner, it will not work. And as I say, for us, DevOps, open source and so on is really to increase the collaboration. We want us, we know that if we target the technology, there is also the people. And these are the people today, you need to change the people and these are more complex. You can buy, you can invest, if you don't change the people and remove people is more complex and remove server sometimes. Don't forget it or do the retirement of application. So, finance cannot be your only criteria to select open source. Can be. We do an assessment all the time between open source and property. It was not the case before. It looked like maybe incredible but when you are a big company you have contract and you need to follow sometimes the contract. My last sentence is a conclusion. It's open source is the only way in the past DevOps and so on to match our contract for our company to deliver quicker and so on. I have seen a lot of news on the open source website and so on about collaboration is open source. Big data is open source. More and more we talk about a topic and we say is open source and there is no other way to go. So, and it's very interesting but when Airbus moves to open source we get with us a lot of company moving to open source of suppliers and so on. So it's also very interesting to have big company moving to open source and to get the other one that follow. If we do it that means there is a sense I hope for our customer. Thanks.