 So welcome everyone to the third Kubernetes Batch and HPC day. Really happy that we managed to organize it again. We have an afternoon full of nice topics and talks with different points of view of what Batch and HPC even means, so it will be quite interesting. So my name is Ricardo with Aldo. I'm chairing this event. And also thanks to the other people that were in the program committee and the talk selection. It's always nice to have interested people in collaborating. So if you feel like helping us out for the next event, feel free to reach out and join the program committee as well. And if you want to submit, of course, we encourage you to submit to the CFP for the next event as well. So this is the third time, as I mentioned, that we do this event. The first one was a year ago in Valencia in May. This was organized. And the main goal of this event was to establish the bridge between what's traditionally used for infrastructure for research and scientific computing and the new technologies coming from cloud native, which were triggered mostly by services. And the idea was to have an event that would be the bridge between these two communities. And it was quite popular. We got like 60 people attending with a lot of live stream attendees as well. And this was the replies we got from the first event, which kind of motivated us to keep going with this kind of events. The second batch in HPC Day occurred later in the year in Detroit, together with Kupkan North America. I wanted to have some numbers for this event, but actually I couldn't find them yesterday. So it was a full day event. So we transitioned from the first half day to a full day in Detroit. And it was very well attended. So that's the reason we do a third one. And I think it's been positive. And it triggered a lot of different actions in different communities. So this time, we do it here in Amsterdam together with Kupkan Europe. So as I mentioned, this is a half day this time. So we listed three main topics when we set up the CFP. The first one was advanced high performance computing. And this is really more traditional HPC environments. But not only. It can also be high throughput environments as well, which we have quite a lot of users interested on. The second one is machine learning and specifically training and what does this mean to run this kind of workload on Kubernetes and Cloud Native. And the third one is data analytics. And we'll have talks that cover a bit of the three as well, both from an end user perspective and from infrastructure as well. So we have five full sessions, three lightning talks. And thanks again, everyone, for the submissions. Keep submitting. I'm sure we'll keep organizing this event in some form. So who are we? We've asked this question before. And basically, in this batch HPC Cloud Native world, we have contributors to the different projects. We have researchers. And we also have end users that are more interested in using the technology. So maybe just as a starter, I'll start. But how would you say that you're an end user? And could you just raise your hand if you consider yourself an end user, just to have a feeling of the audience? All right. And contributed to some projects, a couple of hands. And researcher or more industry, but mostly from the research side of things, right? OK, so we see it's a majority of end users, which is kind of the expectation for this event as well. But the goal is really to collect the requirements so that we pass them to the projects and we push the landscape forward. So the three main groups that people hang out on is the Kubernetes working group patch. And this is where the core work of expanding things like the job API and the scheduler is happening or being led. And Aldo will talk a lot more about that as well. Second one is the CNCF research user group. This was created after Kupcon Barcelona in 2019. And it kind of gathers people from different research communities. And again, it's kind of guides in one entity the requirements from these communities. And then the CNCF patch system initiative. And this is a broader initiative than the working group patch in Kubernetes that looks at other projects also in the CNCF landscape and other parts of the stack. So the goals of the event is to bring together community experts. I mentioned contributors research and end users. And please try to network as much as possible during the event. And finally, the main goal is to make sure that Kubernetes becomes the best tool to manage infrastructure also for these kind of workloads. In addition to the collocated event today, in the main conference, there's a couple of talks that might be interesting for all of us. I tried to have a quick look at the schedule and highlighted tomorrow there's a deep dive into the scheduler in Kubernetes. It's obviously not just about these kind of workloads. But if you're interested in how it all works or even in contributing, do attend. Then there is a gathering from the working group patch in Kubernetes and with an update there. There's also a talk about efficient access to GPU resources, specifically about the different mechanisms you can use for sharing GPUs and optimizing the overall usage. The day after on Thursday, there will be a talk regarding Unicorn and the scheduling framework. And then on Friday, there is a talk about enabling HPC ML workloads with the latest job features in Kubernetes. So all of this, I think, is quite interesting. So it's a nice follow up for this call. Just a quick reminder about the code of conduct from the CNCF to scan for more details and for there will be captioning translated. This is regular standard slides from KubeCon as well. And that's all I will say. And I think we can jump in immediately with the first talk from Dmitry. Thank you.