 studies, open source demos, industry conversations, and all the latest updates from the global community. We are here live on Thursdays at 1500 ETC, streaming on YouTube, LinkedIn, and Facebook. So thanks for everyone that's joining us already. My name is Erin Disney. I run events for the Open Infra Foundation and I will be your host for today. I've actually been producing the show behind the scenes for the last year, so super exciting to be in this chair today. Thanks for having me. Like I mentioned, we are live. I've got my own line of questioning for today's panel, but we'll be answering audience questions throughout the show. So feel free to drop questions, comments, or even hellos and shout outs to the panel into the comment section and we will get to as many as we can. First off, I want to, hang on, one second, here we go. I want to thank all of our member organizations who make the show possible. Huge thanks to their support. And if you're interested in joining the foundation, check out openinfra.dev slash join to learn more. Before we get started, I just want to take a minute and reminisce a little bit about the last time we even saw each other. For some of you, it maybe was the Denver summit in April of 2019. When it snowed, it was not snowpocalypse and Dublin levels, but it still, it was snow at a summit. Hard to forget that one. Or maybe the last time you were at the Shanghai show, which was our last in-person summit in November of 2019, was so great to finally be able to bring the community to China and witness the explosive growth of open source adoption happening there. It was an amazing time. Can't wait to get back there either. So now, after two years, since the last time we saw each other, I'm just so excited to be here talking with today's awesome panel about our first summit back since 2019. We are thrilled to be going back to Berlin. For the first time, different venue. I, it's not going to slide here to show you pictures of the venue because I'm just obsessed with how gorgeous it is. It's, for those of you that remember the last venue, kind of a train ride, but this one is in the center of town. The Berlin Congress Center is the name. We will be in Berlin on June 7th through 9th. We have reg live already, sponsorships are live, and the reason we're here today, the reason for the season, CFP is now live as well. We have a deadline on that for February 9th, so we wanted to get in front of you today and just kind of talk about all, all the things you need to know to be able to submit to the CFP. Also wanted to show this is the fullest of tracks today. We're talking about four today, including a new one new for this summit is hardware enablement. We have somebody here that is on the programming committee to talk a little bit about what we're looking for there. You still have a few weeks left to submit. And like I said, I'm joined today by some open info experts and members of this year's programming committee to help break down what it takes to get your presentation selected and on the official summit schedule. So panel, please join me. Let's chat. Awesome. And okay, also before we get started, fun fact, this is definitely a first in open info history. We have all a names except mine, which Barista's frequently think my name starts with an a so I feel like I may be part of the club. So yes, today's episode is definitely brought to you by the letter a welcome everyone. Thanks for everybody. Awesome. Well, let's kick things off with a round of introductions. We'd love for everyone to sort of tell us a little about yourself, the track that you're representing this cycle and why it's important to you. Arkady, do you mind going first? Certainly. So my name is Arkady Konevsky. I have been around the open infrastructure and open stack before that for the last 10 years and being actually on the open infrastructure board of directors for four years. Awesome. Do you want to tell us a little bit about hardware enablement and what what this track is going to cover now that it's brand new? Absolutely. So I'm extremely excited about having a new track for hardware that have been kind of a pet pet project for me and I have been involved in that since the beginning. And we, you know, the open infrastructure we have been working on the on the bare metal and hardware for quite some time. The ironic project cyborg project bare metal seek have been around for a very long time and have been very successful. But this for the first time we are having a dedicated session for the hardware. So obviously, you know, the first thing first, I mean, somebody need to manage the hardware and how we're going to manage that hardware is the kind of the crux of the matter. And there it comes kind of in two stages. So one is enabling the hardware and what are the tooling, how are we going to do that? And the second one, how do you do the lifecycle? And it includes all the different pieces from the discovering of the stuff, you know, setting up the biases, the firmware, various components and managing them through the lifecycle, you know, how do you want to boot up your images from and so on and so forth. So all of that is part of that thing. And as you go deeper into the into the things as you are wanting to optimize how you're going to use a component tree from the GPU, FPGA, smart needs, you know, all of that is part of that things. And it's kind of related very quickly of how the next things will be had, you know, the next layer will come up, which is networking and the bare metal where you're dividing the hardware for specific use cases. So there are some fair amount of the workloads which are very intricately dependent on the hardware and you want to be able to manage them together. So that's all part of this of this of this topic. So any submission on that will be will be welcome. Obviously, there is a dependency on various standards, which we are relying on to make that happen from the OCP from the Redfish from IPMI before that and so on. So this is very active areas. And there is a lot more work dependent on that. And then how it is utilized by the different classes of the workload from HPC, Telco AI, they're obviously of great interest, not only in design of hardware, but also how we can manage it as a part of the of the problem. So if you have any questions, please post them and we'll be happy to answer them. And I'll pass the button to the next participant. Thank you so much. Excited about that one. Fun to have any track. Next up is getting started. Amy, this is one of my favorites. We actually had an episode. It was a Ben Silverman OpenSack 101 episode that we featured on Open Up for Live over the summer because it's one of our most viewed videos ever. And it's a recording of his CFP submission that he presented in Sydney in 2017. So I love this track. Tell us a little bit about you and what you're looking for here. Yes. So my name is Amy Marish. I'm a Principal Technical Marking Manager at Red Hat. I also serve on the Open Infrastructure Foundation Board of Directors, the Technical Committee, and I'm Chair of the Diversity Inclusion Working Group. And I'm also core on OpenSack Ansible. So I've got my fingers in everywhere. So the getting started track is great because it's a great way to bring people into the communities. And I'm going to say the word beginner level twice here, but it's a beginner level track to learn the basics about all infrastructure-related topics. So things that are important that you can contribute, you know, talks on are ways to contribute to a project. And it's not just code. It's documentation. It's mentoring. So if you were an outreach intern and you want to talk about your experience, this is the topic for you. This is where you're going to put that in. So also how to begin contributing to a project. So a project introduction, you know, what is CADA all about? Onboarding for new community members. I often do getting Garrett Lunch and Learns under this track or separately. But it's a great way for people to take those first steps to joining your project. How to use open-in for technology. So that OpenSack 101, great example. Airship. How do I get my first instance started? So again, very beginner level talks. So keep it simple. Keep it to the beginner level. This is your introduction to people. And I think that's about it on getting started. Awesome. Love that one. Thank you for participating in this one. I think it's just so important, especially as we add new members to the community, just making sure everyone's informed and knows how to get started. So thank you, Amy. Awesome. Next up, Arna, Private and Hybrid Cloud. Definitely one of the most popular tracks historically. Introduce yourself and tell us a little bit about this one. Right. So hello, everyone. My name is Arna Wiebeck. I work at CERN, the European Organization for Nuclear Research. I joined the CERN Cloud team in 2014. I mostly look after operation of our cloud. But I also have contributed upstream to various projects such as Cinder, Manila, and the Core and Ironic. And I also share the OpenSack bare metal SICK since a year or a year and a half. So this track, as you can see, has quite a wide range of different topics. But most of them evolve around the how to do something, how to design, how to pick, how to address certain issues, how to operate and run. And this is probably closest to what I'm doing in my daily job. So this is why it's like where I can probably contribute the most with my expertise. So what we are looking for in this track is probably something or would be a lot around deployments that tell us how they do specific things. So it would be a lot of show and tell of specific deployments, explain how things were set up and why certain decisions have been taken, what kind of hurdles employers had to master, how they dealt with recurring issues, like issues that are, for instance, occur only or occur regularly and are not addressed by the software at the moment or that have issues that come up only once where you had to migrate from one, say, configuration to another. So to give you an example, for instance, something that is that we introduced lately was like to introduce availability zones into our Cinder deployment. So while, of course, the system was in production. So how do we address these kind of issues and how you do this in like recipes and sharing experiences? Probably an interesting part of this talk. Love it. Recipes. It's good. Thank you, Arna. And then finally, Armstrong, thank you so much for being here. Now, this track has been coming up a lot lately. Now, there's a lot of excitement around AI machine learning. So please introduce yourself and tell us a little bit about this track. Hello, everyone. My name is Armstrong Fongem. I'm from Canada. I'm affiliated with the maintenance, construction and intelligence of software laboratory. It's at Kingston, Queen's University, Kingston. So I used machine learning and AI in my research. And the intercession is between the human-centered AI, release engineering, which is in software engineering, and the ecosystems around the deployment process. So I'm affiliated with the release team at OpenStack. And in my past experience, I worked with the Microsoft research team, especially the productivity and intelligence laboratory this summer, where I was asked to do some work on improving their developer's velocity. Previously, I had proposed setting algorithm using machine learning like reinforcement and supervised learning, and also building black boxes for vehicles to reconstruct accident scenarios. So this area, it's something that I'm very passionate about, especially on data science, using machine learning and AI, like swarm robotics and intelligent agents. Most importantly, the area of ethical use of AI, that's like the big thing that I'm very passionate about. Not just building the model, but make it bias free using the ethical way of representing data and explainable AI. So I think on the track, it's very broad range from AI, which is the umbrella theme in artificial intelligence. Well, I will not really go into all the details because everything right now, AI is used everywhere and everybody needs AI. So it's a very broad topic. So try to look at, open it in that dimension to welcome as much as people could represent or mimic human intelligence in a way in the application and being explainable. We also have topics that are in the research areas. For example, scientific research, people have been doing fantastic works that we would like to see in this kind of conferences where they will showcase how it can impact life fairly. We need a high intensity workload for the HPC, the high performance computers and the high throughput computations. These are also fairly recent things that are really coming up and helping AI to get more meaningful for deployment. That said, we'll also go into deep learning, which is the new things in the machine learning to see how we could deploy pipelines right up to production. Natural language processing in the area of businesses, banks and many other hospitals to assist doctors to do diagnosis with their clients, chat boards and voice recognition systems using all this kind of applications that are open for this track, especially running applications in open infrastructure platforms will really be of great interest as well. Notwithstanding, there are innovations that have been happening of recent using the FPGAs, the data processing units and the graphical processing units. All these in large kills are more than welcome. Awesome. Yeah, there's a lot to cover there. I know, like I said, it's just coming up a lot lately. I can see that one being a popular one, this cycle as well. Kurt, thank you for accurately branding this. I should have come up with this myself. You're absolutely right. This is definitely the A team. Well said. Awesome. We also did have a question from the audience. I believe, Arna, this would go to you. What do you think about this as a topic for the public cloud? Right. It's a question about whether a talk about, I read this right, NVMe over fabric would be interesting in a public cloud and a public open cycle. I think that sounds very interesting. I think quite some activities like enabling NVMe over fabric. I think that's an interesting topic. Yeah. Awesome. Keep the questions coming. We do have a couple more things we're going to talk about, but audience questions we will jump to. Questions for everyone? Kind of combining, we talked a little bit about the topics that each one will cover, but what types of submissions would stand out to you or are there any exciting trends within your track that maybe wouldn't have showed up a year ago, but are definitely buzzworthy now? We'd love to hear any comments from each of you on those things. We want to kick it off. Yeah. Maybe I go first then. One of the things that I was thinking about is everything that goes around integrating upcoming technologies or ways of working into existing clouds. The cloud is becoming, at least in my impression, more and more diverse. There's new things that pop up. GPUs are now brand new, but it's still something that needs to be integrated into the cloud. So how do you provide them? How do you configure them? Do you use special GPUs or pass through? Arm is a very hot topic at the moment. So how do you integrate arm servers into your cloud bare metal as well? Using bare metal APIs, do we have this? How do you integrate this? Integration, I think, is one very interesting topic to also maintain your cloud and keep it interesting for your users by adopting these new technologies. Everything that evolves around this I think would be very interesting. Love that. I think one thing coming out of our recent keynotes was Loki. So getting started on how to use Kubernetes and OpenStack and just making sure you do keep it to that beginner level would be a really great fit for getting started. That's great. I'm sure Alison Price is very happy with your Loki plug there. Amy, thanks for that. Awesome. Ercati or Armstrong, any thoughts here? Yeah, I can go. So the things that I would really look at stands out. When we are talking about AI, AI is not new. It's been in the practice in literature for right back in the 60s. We look at interesting works, but there are enabling technology like what my colleague just said. The hardware, the cloud, the data were not there massively to make meaningful decisions. So things like Zip Learning were even forgotten. I remember when people like in Hamilton or many others in the father of AI and Zip Learning, he got frustrated at some point. His provider told him, forget about these things and take something interesting. That is the hot cake today. What made it so interesting data? So most of the things that really stands out, we help people filter out data to make it to remove bias, to make it well represented because we have seen a lot of discrimination happening in the domain of AI. People are kicked out from system just because they don't have the kind of accent that we're used to train those models. So when they give a common other speak, it cannot identify them, push them out. We've seen people, the models that have been trained using some kind of selected male figures when a qualified lady comes, it cannot identify that, classify it as those kind of potential people because it was not trained using a representative structure of women or females. So we really want to see this kind of consideration when data is filtered out, remove every kind of bias that could possibly be and let the submission try to know the limitations of things they are doing, explain it in a way that the prediction will be so much meaningful and trustworthy. Well, yeah, I would definitely go to that talk. That sounds incredibly interesting. Awesome. Arcadi, any trends or topics that would really stand out to you in hardware enablement? Oh, you're muted. Hardware enablement is a brand new, so it's not like we have history of the trend which we can represent on the summit. However, the things which would really stand out to me is that the hardware enablement have been there for a while, but the hardware enablement, optimizing things specific for a specific type of the workloads or specific use cases and the ability for you to manage that in an automated way is the one which should really stand out because you can set up things in a manual way and optimize it any way you want it relatively easy, but to have it done automatically and have it adopted for different use case optimization as you go and especially if you can adopt it such that portions of the cloud are adopted for one thing and you can modify it based upon, automatically based upon the workload which is coming in. So those are big things and this is an area which is very hot and urban for a couple of years, but there is still a lot of the gaps open questions. I think just to add to this what Kari said to support him on this, I think there's a difference between something is supported in the software, like the cyborg is around for a while, there's support for GPUs since a while, but making this work and maintain this over years, keep it running and automate this, I think this is an important aspect of things and as Kari said, there's still a lot of work and there's still a lot of interest in this. Love it. Awesome, well we talked a little bit earlier about obviously getting started as its own track, but we've got first time submitters for all of these tracks. Any tips, advice, bits of wisdom that you would share for folks that maybe have never submitted to a CFP before, ours or any, like I would love to hear some advice for those first timers. Yeah, I think one thing that's really helpful is a catchy title. We want something that's going to immediately bring people in, a good comprehensive abstract, you know, not just what you're going to talk about, but what people are going to get out of coming to your session. I think those are really important. Now we as part of our process, if you submit to the wrong track, we will put you in the right track, so that's important too. So don't panic so much about what track. I mean try to get it right, but you know, if it's not, we're going to try to find your best place. I know we're not talking about the process at all today, but if we do get too similar, we might reach out to you to combine, you know, so don't be so concerned that yours is the most unique thing on the face of the planet. Just make sure it's a good abstract and a good talk. That's great advice. I think like in order to like triage the talks, it's very important that it's understandable. It's comprehensive, because it's like limited to, I forgot, like a thousand words or something. Yeah. Or a thousand characters. I don't know. There's some limits. So like make sure like the program committee can understand what you plan to talk about. So I think that's important. We get a clear picture. And then the second thing I would like to highlight is that, you know, don't underestimate what you have done. And if it's interesting to you, it's probably interesting to others. So don't think like your work is not relevant or not important enough to like have a talk at the summit. I've seen this in the past where people think, oh, that's not relevant. No one would be interested. And it was very interesting topics that a lot of people were interested in. So don't hesitate to submit is probably my main message. It's great advice. Yeah. I think adding to what Amy have said and Annie, the abstract is the key. You really need to put your selling point up front. So like dragging people's curiosity towards listening to your talk. Sometimes people might want to use a lot of high level language, some abstract, something that will even scare people. Keep it simple. Go to that explanation that a common person when I use the word common in the negative sense, somebody who is novice in your area could pick it up and understand what you are talking about. Let that selling point be the take home message. So everybody reading at your abstract will say, oh, definitely I need this. The implication of your work should highlight itself. If you do something, try to see the impact yourself. What are the implications of this work? How will it advance the use of this technology for our common day-to-day uses or for the particular application? Highlight that and really hit hard on your selling point. Thank you. I'll just add that all of those points are exceptional good and it should definitely follow that. The key thing is what is the novelty? What are the things which you expect people to learn from that? Because that's what makes a big difference. Even if you apply the existing technology in a different way or for existing use cases and so on, all of those are great things to show and demonstrate how useful those things are. So the most important things don't be shy. Don't be scared. If you submit it, you will get a feedback. Even if it won't be accepted this time, you will get a very good feedback to help you be productive the next time. Or if it's good enough right now, it will be accepted. Yep, that's a great point. We had a couple of questions. Sorry, was I talking over somebody? I was just going to point out that we also had the lightning talks. Yep. So if you're not confident enough to submit something for a full-length talk, consider doing it as a lightning talk. Yep, that's very true. That's a good option just to add also sometimes it is helpful to look in the past submission. See what people did in the past is the full talk themselves. Just watch some few videos that open infras put up there and try to analyze in your own understanding. What are the key points of these messages? How did the presenter go about presenting their fact? It guide you to be original in your own finding and try to keep it within this kind of range. I love that. Like I mentioned before, that Ben Sogerman video, it's from a Sydney summit. All of our summit videos are out on our YouTube channel, so definitely check those out. Great resource to be able to see what we're looking for, what's gotten accepted in the past. Love that. We did get a question just asking for some clarification. So each individual can submit up to three talks. That is correct. And then Arna, it is a thousand words is the abstract limit. You are correct. That's a great memory there. Just some clarity on some of those things. That's quite important. I remember this because I had written an abstract somewhere else and then I tried to submit it and I was working on it for quite some time like working on the phrasing and then only to find out that I had to shorten it by 50% in order to fit it in. So I think it's good to know up front that this is limited to a thousand words. Yes, very good point. We have another question from the audience. Marcus from Texas asking if it helps to include links to past talks. Any comments there? Yeah, I mean, it definitely can. It shows a little bit of commitment that you are going to follow through the process and that you will give your talk because there have been people who have in the past got submitted, got approved, and then didn't show up for their talks and things. So that does give you a bit of history that the person will talk and their level of talk, especially when you're like on the fence of whether it may be too technical for your track or more advanced or does this really fit in? And you can look up somebody's history and see what that abstract was and how they actually presented it and whether it will be a good fit or not. The submission form has a specific part for this, I don't remember correctly. It does. Yep, and apparently I said a thousand words. It's a thousand characters. So just making myself as well when I said a thousand words sounded a lot. This is my first time host problems. So yes, a thousand characters. Thanks to my team for clarifying that. Awesome. So we do have a slide. We kind of spoke to this a little bit earlier, but I wanted to at least run through since all of you had some great things to add here. You know, tips and tricks are great, but like do's and don'ts, I love some of these. Anyone feel like calling out some of the ones that you added here? Sure. I mean, I'll go. So as usual, when you're presenting, make it interesting to as large audiences as you can. I mean, it also will be of interest when the program committee looks at the abstract, the larger audience is impacted by that, the better it is. Demo is always a big seller. People would like to see the demo because it's the most interesting part. So this is always in the case. Don't try to do sales speech. I mean, this is absolutely no, no. I mean, there is a place to do that, but not part of the program presentations. And we have a separate place at a forum where that is applicable, but not for the submissions. Lesson learned. What did you learn from doing the work? And, you know, how do you impact what kind of things you're planning to do from that going forward? And since we are the open infrastructure, collaborating with users and developers is one of the key aspects of how we operate. So, you know, demonstrating, you know, as part of that, or maybe joint submissions are always very welcome. Yeah. Okay. One of the interesting areas in AI machine learning, as I said earlier, is cleaning your data. So one thing people in the AI track really wants to do is to remove all kinds of bias, ethical, unacceptable behavior that the data may represent. There are tons of those. There are tools out there, there are a lot of mechanisms or techniques that could help folks to get in touch and will keep our contacts to give pointers and integrate us how to do this kind of thing. Try to keep your work simple and original. Don't try to solve all the world's problem in one submission. Break it down to one meaningful, impactful model, which can easily explain by yourself. If you have a lot of things on your plate, you get confused and you start going around. Then don't over-assume the audience understands all the high-level things or just try to keep it simple. And don't come and be criticizing the works of other people or this guy, try to add value of something that have been built upon. Even if in the past they did some error, they did some great work up to the point where you are adding value. So pick it up from there and add value. You will keep your work so interesting. Then don't over-exaggerate your result. This is commonly practiced in the world. When people are abusing statistics, they will tell you, oh, I have a 95% accuracy. I have 80%. Those numbers themselves don't tell any good story. If you don't know the population you are talking about, we don't know the size of things you did, you just come with some 95% of the time. Please just watch out those kind of minor things. They can mislead and lead easily to rejection if you don't really explain a lot of things around your data. So you really want not to over-exaggerate or to blow off your implications or impact of your result. Keep it reasonable. The simple ways are always the good ways and the best, surprisingly. That is great advice. Anything else you guys want to call out? Yeah, there's one thing. It's kind of valid for demos and for lesson learned. Don't be afraid of failure. When lessons learn, talk about things that didn't go right. And then what you learned and how you did get it to work in the end. A demo, if it blows up in the middle, think of what everybody's going to learn watching you fix it during your demo. So failure is a great thing. Don't be afraid of it. I love that. Arna, did you have something to add? Yeah, I would like to add something to make assumptions about the audience. Many of us are working with OpenStack since several years. But nonetheless, I don't think that we are experts in all areas. People are focusing on the storage part or the networking part or specific areas within OpenStack. So don't make assumptions that the audience knows everything about the topic you're talking about and start with something right in the middle. I appreciate if there's some kind of introduction that helps the audience to understand the context and where this is going so that you have at least initially make the audience a little bit comfortable in the first third or so so that people hear something that they're familiar with immediately confronted with something that is overwhelming and you lose the audience right away. I think that's also one of the lines what Armstrong said, like keep it simple to the point and at least initially try to take the audience with you is very important. And then also what Amy said, I like very much failure. I still remember presentations where there was a demo and something failed. I still remember what the demo was about while others I have seen were just worked. I have forgotten. So it's very good. I mean, it's not only about praising and how brilliant someone is. I mean, it's also good to call out failures or deficiencies. I think that's also quite interesting. Yeah, that's great advice. Actually, I remember some episodes where talking about the failures is just as therapeutic as hearing what went well just because you can relate to some of those things. I think that's great. Awesome. Very good. Well, we don't have any more audience questions. Do appreciate those that have come in. Before we wrap up, I think, you know, is there anything else you want to add? I know we're definitely going to be posting everyone's how to get a hold of you. We're still the truck chair or the programming committee deadline was yesterday. So we're still getting all of those notifications out. And we'll definitely be doing some office hours, that sort of thing. So as you're working through your CFP submissions, you'll be able to reach out and speak with the programming committee. I think that's something that's so great about this program is that everyone makes themselves accessible and available for questions to help along the way. So we do have one audience question that came in and then we'll post everyone's contact information. Do you have to be there physically to present or can you present virtually? Great question, Sarah. Right now, we are not planning on a virtual element. So it would need to be a physical presentation this round. But great question. But let me just add in that there is a really great travel support program. There is. Thank you, Amy. Great point. Those nominations or that form to submit is live. We'll be posting, we've got the, hang on, one second. Looking through my banner, seeing if I've got one I can show here. We'll have the summit website up here in a second that you can submit for that here. Let me just show it right now. So you can learn more about the summit at the openinford.dev. The travel support link is there. Definitely submit for CFP, I would say. And then submit also for TSP and we will do our best to get you there if the travel is a concern. Awesome. I believe we've got some banners up. If everyone is comfortable, I'm going to show how to get a hold of each of you. If people do have questions, particular to each of the tracks that you're representing, let me grab these real fast. Armstrong, starting with you. You've got your IRC and your email here. Thank you so much for providing those. So you want to get in touch with Armstrong. Here's that information. Arcadi, thank you for providing your email address. Leave that up for a second for folks. You've got Amy's IRC and email. Double email, way to go. Now you're winning. And then Arna, awesome. Thank you so much for writing your IRC. Sweet. We did have one more audience question and then we'll probably close this out. Not sure who wants to take this. What's the best way to decide on the complexity of a topic? The interface invites to determine when submitting the abstract. Any advice on this? It's hard and I'll admit it's hard because if you're very comfortable with a subject and you're like, I can do this in five seconds, it's beginner. And then you read through it and then it's not. And if we do run into something like that, I believe we can change it in the system or we'll contact you and say, we've read through your abstract. We really think this is an intermediate level, not an advanced level. Are you okay with us changing that? Again, you're not on your own with this if we need to reach out to you to get clarification and make sure you're correct, we will. Yeah. One quick way of getting around this kind of breaking down complexity. If you are around somebody who is completely ignorant about your topic, just plead for two minutes and just represent your ideas to that person. If that person like, what are you talking about? Okay, you cannot try to explain it in a way that somebody with complete novice in your area could at least speak of one concept and tell it back to you, this is what I got. Then you are communicating from that point. That's a great answer. And great question. Thank you to the audience for asking some questions today. And thank you to the panel. Appreciate you all being here and sharing all of your awesome insights. I think this was super helpful. Yes, this is the link. This is the link you need, CFP.openinfra.dev. Deadline is February 9th. Let me pull up that slide one more time that's got everything you need to know. Reg is live. Sponsorships are live. CFP is live. Everything is live. We're in full Berlin mode. Definitely get those talks in. Like I said, we'll be publishing some office hours for the programming committee coming up in the near future. We also have, oh, here, sorry, one more time. Here's all the tracks just for your reference in case you missed this slide at the beginning. All of this is out on the CFP tool at the link that was just provided. We also have another episode of open for live coming up on February 3rd, also at 1500 UTC. This will actually be another in our large-scale open stack series, which we've been doing for the last year, every month or two, where we feature operators discussing both their deployments and operations. This episode will feature OVH cloud, so make sure to subscribe to your preferred platform so you don't miss it. Also, remember that if you have ideas for the show, we definitely want to hear from you. Submit your ideas at ideas.open impra.live, and maybe you'll be on a future episode. Thanks again to today's panel. Remember to get your CFP submissions in by February 9th. Mark your calendars for February 3rd for that upcoming episode, and we will see you all on the next episode of open impra live.