 So, I'm here to talk about a revolution, but first, a little bit of context. My name is Jan, and I don't understand why it's going so fast. I've been working at SNCF for six years now, and I co-started the Edge Initiative there and SNCF. Well it's a major transportation company here in France, and specifically trains, and to give you the idea of the scale of the impact of our Edge project, SNCF operates 15 southern trains every day and transports 10 million people, which is about 15% of the French population. So why make a revolution on such a critical workload? Well, because we need it. Oh, yeah. There seems to be an issue here with my poor point. I'm sorry. Let's go. Can I do that? Okay. Sorry. Yeah. Let's try that way. Okay. No. It doesn't seem to want to work. Sorry. So, yeah, trains have 40 years long life cycle. They are industrial environments. And as such, every technology that is embarked on trains has to live that long. For IT, it means that we have software functions, secure ones like brake controls or door controls, and comfort functions like the custom Wi-Fi. And until now, from an IT perspective, every function is managed the same way, which is a specific software on a specific electronic board with a specific operating system racked in a unit such as this one, which means we have a very limited physical capacity. And since it does not support remote deployment, we have to have onsite deployment for every function or for every update of one. Okay. And due to that, we have a two years long time to market for every function or every update to have a complete deployment. And that's why with the IT, the embedded IT teams, we decided to build an edge system to try to solve all these issues. So how does it work? First, to not reinvent the wheel, we tried to build on existing capacities of the trains. So we have 40 communication. And we had stored out hardware, at least hardware that is certified for trains that we can use to build as a shared capacity provider for our functions. Then we built a custom stack for this hardware with K3s, a custom operating system, which is a bit like Keros, but self-built, self-developed, and a VPN client to secure communications and help us be network agnostic. And these 4G communications brings us in the SNCF network. And there we built a VPN proxy to allow us to deploy a launcher in the cloud where we have scaling capacities, and to use FLIT as a manager to make sure that we deploy everything we need on the trains. To help us have a major availability on trains, we also put a registry inside the VPN. And as a bonus, finally, a bit like Eric said, we built a self-custom controller to help us have automatic registration and configuration on everything that is not managed by FLIT, no launcher. And where are we on this revolution? Well, yeah, we're going there, but it's taking a long time. I'm sorry. Yeah, it doesn't want to go on my time. After a one-year POC and one-year project to deploy everything, we are now production ready. And we are waiting for train maintenance to allow us to deploy our hardware on the trains. And it takes us at least two years to do that. So what are we doing in the meantime? We're working on redundancy, function redundancy, and we try to develop and migrate existing functions on the edge. And we try to certify the project to be able to host secure functions on the edge system. And also, we've started to work on the Harbor satellite with the Harbor team, which is an edge-minded solution to help us ensure image availability on the edge. So thank you for listening to me. And feel free to contact us if you want further details on what we do and have an amazing KubeCo on everyone.