1:34 Those downtime worries goodbye, my friend. Quality and uptime, the stacks will descend with a platform that Hello, and welcome back to the Rawkode Academy. I'm your host, David Flanagan, also known across the Internet as Rawkode. Today is another episode of Rawkode live, the show where we get hands on keyboards and take a live look at cloud native and cloud native adjacent technologies. And the title of today's episode is DevOps on Autopilot, which is a bold claim, so I'm very excited to see Qovery and be joined by the CEO, Romeric. Hey. How's it going? Hey. I'm good, David. Thanks for having me.

2:46 Introduction & Welcome

3:20 How are My pleasure. Thank you for joining me. That's an amazing song like that you get, like, for the intro. It's crazy that you can do now with Jet AI. Yeah. I started doing that, I think, like, four or five episodes ago. I mean, AI tools are getting so good now and generating music is I mean, I I I generally enjoy them. I I I think they're little catchy numbers. I mean, they're not, you know, pushing the envelope or anything like that, but music doesn't always have to. So I think it's nice that we get a special little,

3:53 you know, song to commemorate our time together today. So That's amazing. That's amazing. Especially that I was, like, doing actually, so it's completely out of topic, but I did practice. I did, like, some music, produce music, and you know how much time it takes just to get that. That's crazy. Like, today, you can just, like, in a few clicks, few prompt, just get something that is ready to be uploaded. Crazy. Yeah. One prompt that are rigos over and over again. I tweak it for each tool and and and each guest, and I'd like the music

4:22 to be somewhat familiar. So it's all kind of similar, you know, mellow, electronic, chilled beat, but it works really well. And, also, it's really good at metal. I don't wanna, like, just deviate right on the music path in the AI chat already, but, yeah, you could do some pretty powerful things with that. Anyway, people are here to learn about DevOps and Autopilot. They wanna learn about your tool, Qovery, and they wanna learn about you. So could you do us a favor and tell us the history and a little bit of flavor of Ruric? Yeah. Absolutely.

4:37 Guest Intro & Qovery Overview

4:48 So just quickly introduce me introducing myself. So I'm Omar, IT, and cofounder of Qovery. I used to be a software engineer, DevOps engineer, like, for a long time. Now I can no longer say that because I've been doing that, like, for years now. I'm still practicing. I still have, like, some, like, open source project that I'm just getting my hand on. But, anyway so, yeah, basically, what we do at Qovery is that we we try just to make sure that developers are much more, like, autonomous in their day to day journey. You know, that they are not the most, like, proficient

5:20 in infrastructure management, like, deploying the applications. All of that is something still under by mostly DevOps engineers or actually so or at least, like, people that they are familiar with infrastructure. And what we do provide is the platforms that helps, like, any developer to deploy their application in their own cloud account. And that's something that is quite interesting because I'm pretty sure you're familiar with Fly and, like, all the other paths, so platform as a service. And, actually, you can see Qovery as a platform as a service, but next generations. Why? Because, actually, you can run Qovery on your

5:56 infrastructure, and Qovery runs on top of Kubernetes. And this is why it's quite interesting because you get the benefits of those PaaS platform, but on top of your infrastructure, basically. So in just a few minutes, you can start getting a fully state of the art infrastructure ready on your cloud account without any knowledge. So we have customer in five minutes. They can do that. And then you can start deploying and shipping your your application. And we have really, like, focused on the developer experience, what makes, like, our platform loved by our users. So, yeah, that's

6:30 we can dig into all of all of that, obviously. Yeah. I mean, back in bold claims, which I love, you know, there's a lot of challenges in, you know, the DevOps space, the platform engineering space, all of this these days. You know, come software architectures are getting more complicated and the with the aim of simplifying, know, developers' time, like work on microservices. But all that complexity doesn't just disappear. It's getting pushed down that stack to the platform, the Kubernetes, the cloud providers, etcetera. And then there's there's just a lot of tooling needed in a space that people can

7:04 still go to sleep at night and not have to worry about the pager going off and a whole bunch of other things. And I feel like we're pushing a lot of this away from developers too and fragmenting things. You know, the dev ops teams, the platform teams are taking on all these responsibilities, but there's not always that collaboration that dev encourages us or wants us to kind of achieve. So yeah. It's interesting times these days for sure. Yeah. Absolutely. And, actually, that's quite interesting what you're saying because we are seeing also, like, when we started the company,

7:33 even Kubernetes were were not, like, the standard. Like, if we were looking if we are looking, like, four years back, like, only companies that they were, like, quite a bleeding edge were, like, considering and starting using Kubernetes for real in productions, and now you have So now it's the it's the standard, like, the operating system of the cloud. But what we are seeing now is also, like, even DevOps engineers that they are proficient in Kubernetes that they are looking to solutions just, for the data operations, obviously, but also to bring, like, a better experience, like, to

8:04 the end user because it's quite whole, like, doing cube control, comments. We have done all that. That's that's pretty good. But in terms of experience, especially for developers, it's not. So Yeah. Definitely. Hi, Ravi. Thank you for joining us and saying hello in the chat. Feel free, anyone else watching, to say hello and let us know where you're from. Rawkode, before we talk about the problems that Qovery solves and get hands on with it and show people how to get started, maybe you could give us a better information on the the origin story. Like, did you just wake up one

8:31 Qovery Origin Story

8:38 day and go and go to fix DevOps, or were you did you have challenges within your own career or your own job? Like, I mean, how did you go? Yeah. I do think it's a classic story. Yeah. I used to start when I start working, and this is where I met my cofounder actually, like, in the HFT, so high frequency trading. We were, like, managing thousands of servers, like, globally. We have a very small team, so quite proficient in infrastructure management. I used to work with VMware system, Solaris. So you remember at the time? So before that, the cloud was

9:12 something. And then I transitioned, actually, is, like, from the infrastructure side to software engineering side. And I was in the position that, okay. Now I was no longer, like, the ones that were, like, dealing with tickets coming from the software engineering department and provisioning, like, infrastructure, but I was the one that were claiming for, like, that infrastructure because I needed just to run my software to deploy live, obviously, in productions, iterate, like, in the development servers and so on and so forth. And this is where, like, I just realized that that's crazy. Like, the experience and when you're

9:47 a developer, it's very it's kind of broken, like, kind of broken. So I used to move, like, from one company to another, like and instead of doing, like, the same thing, which was, like, at some points, building a platform internally, like, into the company that I was working for, I decided just to say, okay. Instead of, like, working for the next successful companies, like, that they they they would hire us and ask us just to build up, like, a platform. Okay. Let's build up a product. Let's deliver that for you to all the companies that would hire us, and that's the

10:20 way we started just basic recovery. That's that's very, like, very classic. You were doing that before. You have done that, like, four years. Now you have the experience, then, okay, let's speed up a product. But it's quite challenging because I'm technical, and now I'm much more, like, in position where I'm I'm a salesman. I'm spending all my day just, like, selling the software, talking to engineer, to CTO, like VPs, and it's quite interesting, like, the two aspect. But, anyway, it's completely something else. Nice. What would be you know, if if people watch us, what's the

10:54 Primary Use Case & Problem Solved

10:59 the one use case right off the bat that anyone's setting right now is struggling trying to wrangle platforms, DevOps, Kubernetes clusters? Like, what is that one first use case that you think Qovery can come in and help people with? Yeah. Yeah. That's a very good question. We have a very good use case, and, actually, it's part of our go to market strategies since we are accelerating and we're getting more and more customers. We do focus and need to focus on one main use case, but it's basically, like, all companies, like SMBs companies that they are transitioning

11:30 from past platform. They need, at some point, they are, like, limit limited by, like, the platform as a service. You are taking hair or cure, render, fly. You get the limits because of technical limits, but also in terms of security, in terms of compliance, and you need to get into the, yes, like AWS, GCP, Azure, scale away. You can name it. And, basically, here, you have two options. The first is hiring, like, a DevOps engineer. When you are, like, a very well funded and famous company and you want to build up anyway, like, a a

12:04 DevOps engineering team, that there is no problem for that. But when you are not an not all companies, they can just afford, like, hiring a DevOps engineering team. Then this is why it's quite interesting because Qovery can be a good very, very strong and good option. Because in just a few minutes, you get a state of the art infrastructure with a very good developer experience. So meaning that you can just stop right away deploying your vision. That's a very good use case, that we are solving with our product, and that's the main use case because it's production

12:33 infrastructure mostly. Yeah. So Okay. One more question, and then we'll share my screen, and we'll actually show people how to get started. Obviously, this is 2024. Right? I think the last twelve to eighteen months have been a turbulent time for open source projects. He's I I'm not gonna name them. I'm not gonna talk about licenses or any of that. Right? But well, very does it is open source. Right? So, like, why did you go down that route? That's quite interesting, and I really and I'm gonna respond to that question. Actually, we don't even define ourself, like, as an open source

12:38 Open Source Philosophy

13:12 company. But if you look at our GitHub repository, we have a lot of and, all deployment engine is, like, fully open source. And we have a lot of components that they open to. But you see, we are not defining ourself as an open source company. Why? Because, actually, the product can't work, like, without the control plane. So we did, like, take the decisions off, and I did take the decision of delivering as much as we can, like, in open source. So mostly, like, the parts that they are installed on the infrastructure of our customers because it's for, like, security, like, reason

13:50 and retransparency because all those elements, there is a lot of small elements that they are running and operate they are operated, like, remotely on their infrastructure, on their Kubernetes cluster, on their cloud account. And so for transparency, we have decided just to open source everything, and we have also, like, a few other elements, like, tooling that we are using internally that they are completely open source and they can be used by our customers. So we have the open source philosophy at her, and we do contribute. We are even doing some contribution. But we don't define recovery as, like, a

14:27 100%, like, a % open source product because, actually, the control plane is, like, not open source today. Yeah. Well, yeah. I mean, I think, you know, I jumped on to the organization on GitHub. There's a lot of open source happening, but, you know, that can only happen if companies can be sustainably making things open source. Yeah. I think that's the way to do it. I mean, it's a very big challenge right now to Yep. Write all these amazing products, make them open source. But then, you know, if you can't pay your bills, feed your family, then what's the point? So

15:02 yeah. Yeah. That makes lot of sense. But I I like the the trust aspect and the transparency aspect. I think it is a really good reason for companies to push out more open source components even if the whole thing has an open source because it does give people trust that, you know Yeah. And yeah. Maybe the same as yeah. I'm sorry, but I'm a I'm a really hard believer in in open source. I do think I'm doing this job because of hope open source and a lot, and open source is really something. But I do think and that's another discussion, but

15:39 there is a kind, like, of I would not say a lie, like, in the industry, but it's really for me, something that is open to us is really open to us. Like, really, you can do, you can take it, you can change it, you can commercialize it, you can do whatever you want, and it must be, like, very important. Otherwise, like, all the software that they are not, like, in that aspect, you can really and because of the license aspect and so, that for me, they are not just they are not really open, but it's another, again,

16:06 like, discussion. And and I I prefer just to be super transparent and very clear to our users as well. Yes. We have components that they are open source, but cover itself is not open source. Right. Yeah. Alright. Awesome. So shall we show people how to get started and what Qovery does? Yes. Let's do it. Let's let's do it. Alright. I am on the Qovery website. As always, on Rawkode Live, nothing is planned up front. I am in a very nice position where I've had my eye on Qovery for a while, but I've never actually kicked the tires on

16:23 Getting Started: Hands-on Setup

16:49 it. So this is this is legit for me. This is my first time ever having to play with it. So we'll go straight to the get and start to guide, and we can see there's a this is quite nice, actually. You've got a whole bunch of, like, tracks and scenarios and tutorials. Wow. I don't often see that. That's awesome. Very cool. I'm assuming we're gonna start with that's a nice effect too. So I'm just gonna geek out over your website. Nice. I'm assuming we're gonna start with Fumble World. So Yeah. And the documentation is open source. We

17:20 open source. Alright. So to begin, I need to get hub account checked. I can manage that at least. And we are gonna sign up for the Qovery web interface. So let's do that. Yeah. You will be you will be surprised by the the the sign up workflow. Let's let's do it. Alright. So I need to pick my cloud. I have access to to to all of them, so we'll just go with AWS. Where's where's DevOps on autopilot? Why is that not an option? That's a that's a good one. That's that's a feedback that we should

18:19 Alright. And now I just need to confirm my email address. I'll do that over here. Yeah. And you see yeah. Absolutely. Actually, you you would be, like, surprised. There's something. So when you sign up, you you you you're blocked by this this page. So you can confirm, and, usually, what you get is that an automated message, Jackie, like, in your mailbox. And then from that, we give you access to the platform so when you reply to it. But I'm gonna just give you so I can give you access, actually. So Oh, okay. I'm looking for you yourself.

19:20 Ah, yes. I can see you. It's good. An enterprise license, please. You have to choose your plan then right after this screen. If you go back now and you and and you go and you move forward and you validate. Yeah. You have Alright. Rawkode Academy. I'll call this platform. Alright. You can start with the free. Yeah. There is no yeah. No problem. Or you can take the team plan, then you have no restriction, actually. It's there is no credit card needed, so that's good. Alright. So you got three plans. Yeah. And up to twenty five years, so

20:02 it's three cluster of 20. So okay. I mean, this is perfect. This is fine for me. Yeah. Absolutely. I like it when it's a a pretty generous free tier. Yeah. Alright. I've got Qovery. Oh. Yeah. You're ready to go. So I let you just discovering the product that's quite interesting, actually. I don't want to, like, influence you all. Well It's kind of a product session. I'm just looking at how you interact with the product. Wait. Obviously, my eyes are already getting right to this big create custom button, which I know I should be back and

20:39 follow on the tutorial, but I I'm just gonna go rogue and, like Yeah. Let's see. So You are doing, like, a 99% of all the users, like, share the documentation. And, actually, it's the it's the well so that we consider product is that if you have to look at the documentation to use it, there is a problem. But sometime, it's necessary still. Okay. So, I mean, sensibly, I I mean, I will do a local machine because I want to see that workflow. But I clicked on AWS just to see what the options were. And we've got our Qovery managed

21:11 cluster or self managed cluster. So I just wanna make sure I understand this correctly. If I I say I've got my own company or my own team, we use AWS, you can I'm assuming if I give you the right credentials, you're gonna manage a Kubernetes cluster for us. But if I have an existing, say, AKS cluster, I'm assuming we just have self managed and we deploy some sort of Qovery agent to it. I mean, is that the right assumption? That's perfect. That's exactly that. You get it right. Perfect. With Qovery managed, I mean, is that in my own AWS

21:46 account? Do you run it in your own account? How does that work? Okay. Yeah. That's a very good question. And, actually, so you do just, like, provide the region the way you wants to deploy. So you have that. The credential so you provide your credentials. So we give the permissions that we need just to create the different resources. But then from that, you have few other options like Carpenter. You can install, like, some add ons on your Kubernetes customs and so forth. We'll generate so at the end, like, the work configuration. So we'll create the VPC security

22:20 group load balancer, the the the cluster in, like, multiple availability zone. So we make sure that you get, like, a really fully fledged, like, production infrastructure running on your cloud account. So it's really running on your cloud again. And the beauty here is that you don't have to invest, like, in day two operation. Meaning, the upgrades of those clusters, but because with AWS, you have an upgrade of EKS, like, every three months. So you need to spend time, like, on that. And that's a big value that you get out of discovery manage, offer because you do once it's done,

22:53 you get the power of Kubernetes, but we are the asset of maintaining that. Okay. Got it. Let's keep things simple. And oh, here we go. So this is where it's become a a common theme recently where people ask me to install command line applications, and then I have to tell them that I use Nexus. So is it a next package? No. Is it a static binary? I do think there is a next I do think there is a next package, but it's not maintained by I'm not sure if it's managed and still maintained by the team. I I do know that it

23:02 Installing CLI & Local Environment

23:34 was a customer that did provide that, but you can find it on the QWERADI. I'm gonna just look at that. Maybe I mean, I can always jump into a desktop box. It's not a big deal. Oh, we don't yeah. But it has not been. Yeah. Okay. I found it. Is 092.8, is that old or recent? No. It's old one, actually. Okay. Let's not do that. Yeah. So let's pop this open. Yeah. And let's see if it's a static. Is it running go? Yeah. It's go. Oh, it should be okay. Yeah. The the path the path maybe is not

24:25 the right one. Like, the Yeah. You can change the path. Like I was looking for local the destination. Yeah. Just dot. Okay. Yep. Makes sense. Alright. I need to trim audit. Yeah. That's a good one. It's not static. Oh, it's not it's not static. It's linked? Okay. Yeah. Dynamic is linked. That's Alright. That's a yeah. He seems to be used to Yeah. I'm I'm getting so used to it. I mean, Nexus has a lot of power, but a lot of paper cuts all the time, especially if it's not in next packages or if it's not up to date in

25:21 next packages, it it it could be pretty painful. But we'll get that extra box just it's only gonna take a minute, and then we'll install it and just run from there. Let's see if there's anything I can do in the meantime. Do I need a Kubernetes cluster locally? Actually, Qovery is gonna, like, spin up everything for you. You just need the docker runtime, and that's it. Okay. So yeah. And really, I should just start writing next shells for all of these streams where I get, like, a proper Linux file system and a directory just for stuff like this. Because, again, this

26:06 this is not the first thing this happens. It's it's it's a common theme there. Yeah. But it's good. It's good. It's, like, re unscript. It's it's live. Alright. Wish I'd done arch instead of the door now. Okay. I I used to I used to be a Debbie Allen guy for a long time. It's been a while now. Yeah. Me too. Until Arch came along and but I just rolling releases are the best way. I always need the the latest and greatest versions of packages, which Nextos is generally really good for actually, especially these days. But there are always packages you find are

27:03 just that little bit behind where they need to be. Mhmm. Alright. So I might need to install the Docker CLI, but that's okay. Let's get Qovery. Again. Qovery off. Alright. We're cooking. Yeah. Perfect. That's good. You even have like, in the case that you don't have a brother, you can just type, like, dash dash headless, and you get, like, a code that you can just validate. Like Oh, no. DustroBox is the coolest thing I've seen in a while. Like, I am in a container. It is a fedora container, but it hooks up everything to your host. Like,

27:49 I'm actually still in my own directory and stuff like that. That's really nice. Yeah. Yeah. It's really slick the way they've done it. Yeah. So great. Alright. You're authenticated now. You can just, like, basically run ready for for this command. So we're gonna need a a Docker engine. Right? Now does it need a Docker CLI, or does it use a Go SDK? I do think it's a bash script, so we are using just the the binary. That's possible that we're using the binary. And you see It's installing all the tools it needs? Oh, no. Yeah. Yeah. I wanted Docker

28:24 though. It's you don't. It's okay. I can get Docker. Yeah. Yeah. Yeah. Absolutely. But, yeah, basically, this command is here just to install, like, we do install what what did you install? I do think a Debian with, like, a k k three s that is running, and then we install, like, the different charts, the QWERADE. We do generate the M charts, and we install the M chart on it. Yeah. Let's see if it works. At least on my side, it works. Oh, we'll get there. The the challenge is is the Internet in my office has been, like, potato net for

29:13 a while now. Oh, yeah. They they really need to fix it. Alright. It's getting there. Okay. So Qovery demo app. Yeah. Oh, there we go. This is all in case three d. Okay. That's Installing how okay. So it's installing all the tools, but it does expect Docker to be on the host, which is fair enough. Right? I mean, everybody has Podman, Docker, or something like that running these days. Yeah. But this this command can be, like, flaky sometimes because we do not control all the environment, like, on your machines. So sometimes we have, like, some weird issues. But,

29:55 yeah, in general, it works. But it happens that it didn't work. Uh-huh. So people, they fall back, like, on the classic installer of running that on their cloud account. Or if they have already a Kubernetes cluster, so they install Qovery on it, and it just works. So yeah. But that's quite convenient here because you can just run that command. Once you've you're done, you can just drop, like, destroy run the same command and destroy, like, the the configuration. Nice. Yeah. Alright. So it now appears to be installing Qovery to the cluster. Yeah. Congratulations. I'll tell you what. I'll do you a

30:52 favor after this stream. I will update the next package for you to the latest release. Oh, if you're doing that, then you will get, for sure, a t shirt or something. Oh, yeah. I'll take it. Don't you worry. That's good. Hopefully, it's getting there. I mean, what's that installed into the cluster right now? Like, what what is that Qovery Home Chart doing here? It's installing. So QWER is so it's written in in Rust. Okay? So we are installing, like, basically, the binaries. I can even share my screen, like, during the installation so I can Yeah. Go for

31:17 Understanding Qovery Architecture

31:33 it. How how things work. So you will see it's, like, public. It's quite it's quite interesting just to get into that. Doesn't support share. Mac OS is just There you go. It's it's there. Oh, it's mine. Okay. Perfect. Yeah. So, yeah, basically, we so how it works, it it works this way. So you have the control pane running on our side. Okay? And you have the the Qovery binaries that they're running on your Kubernetes cluster. So what we do in style is that we create a namespace, Qovery namespace, and we do install those binaries. Those binaries,

32:09 they are they are written in Rust. Okay? So that's the part. Like, all those components, they are like open source. So you can dig into that, and you can see exactly how it works. But, basically, the way it works, you will see it's very dumb, but it's done in very good way. But we interact with the Kubernetes API, but we also do generate, like, all the configuration file on the fly. So based on the configuration and the stuff that you select on the interface, and then they are translating into the control plane. Okay. And you see

32:39 something which is quite interesting is that our control plane is like a public API. Okay? And your machine is not exposed. Right? So you are installing QM recovery on your own Kubernetes, like local machine. And what happened is that it's the cover engine that is getting the connections to the control plane and not the opposite. So which is like a reverse connection. And then from that, the cover engine your cover engine, when you deploy an application and you do some operation, actually, they pull, like, task from the control plane. And then they are executed, and then

33:17 they report back to, basically, the control plane so you can get the information from the interface. Alright. Nice. Yeah. Those are pretty cool diagrams. How'd you make those? It's Excalidraw, of course. Scalidraw. Nice. Yeah. But now I'm using t l draw. Now I don't know why I'm yeah. I think we've got t l draw now. I see. I like it's ScallyDraw because I've got a Versus Code plugin, so I can just pop up like a new window in this. I mean, I'm sure TLDraw may have one too. We now have Qovery clusters, so I am gonna switch this back now.

33:54 Yep. Ta da. So I'm assuming So now yeah. Hopefully, if I come here And you go back. We have a cluster. Alright. Let's click. I'm not even gonna look at the tutorial. I'm just gonna start clicking on buttons. But alright, I've got question there. Clicking yeah. I'll let you click everywhere if you want. There is no problem. I'll let you discovering the product and what you have to so, yeah, what you have to do. But, basically, here so maybe I can just give some explanations, but you are getting into the settings of your cluster. So we have a

34:31 concept, what we call advanced settings. So we'll not cover that. But, basically, once you have your cluster that is ready, okay, you can start deploying application. Okay. And you see something very cool is that, for instance, in my demo account, I have, like, dozens of clusters. So you can have clusters coming from anywhere, AWS, GCP, Scaleway, Azure, and you can plug them and unify the work experience. That's why we have customers actually that they're using Qovery. So, yeah, basically, yeah, you have to click here. It's not super, so we are reworking on that interface. But, yeah, that's exactly what you

35:09 need to do, actually. Alright. I'm gonna call it production. Yep. I know it says I've got constraints, but I wanna know what those are. So I'm just gonna go straight and Yolo it. Yep. I'm gonna click on settings first just because I'm curious. Uh-huh. Alright. Very nice. Yeah. Alright. Yeah. I'm I'm starting to get a feel for this now. Cool. What can I deploy? Let me see. Yeah. Nothing complicated. Nothing complicated? So you can deploy an application coming from, a GitHub rep. So you can you can take do you have some codes, like, on your GitHub repository,

35:50 Creating Environment & Deploying Service

35:55 for instance? Sorry. Can you say that again? Yeah. Do you have any applications available on your GitHub repository on GitHub account? You mean, like, a deployment dot YAML? Yeah. Or you can deploy oh, you can deploy your hand charts if you want. We have a very nice experience for that. You can deploy your container directly from a container registry. You can deploy an applications coming from GitHub. So Qovery will clone the repository for you, build if you have a Docker file and so on and so forth. So you can really choose. So just put yourself, like, in the shoes

36:30 of a developer. Okay? You are software engineer. You are used to use a pass, and you want to deploy in applications coming from your Git repository. Yeah. Okay. Container registry. New registry. Get go away with password. Anonymous is fine. And image name, oh, Rawkode. Mustard v one. I'm gonna just check back. Yeah. That's a that's a that's a good thing just to to take a look at. Yeah. Here we go. Just V one and v two. And and okay. What's the what's the the image name? Like, because it's longer than just that, I do think. Maybe it's the

37:32 yeah. Oh, you find the tags as well. Yeah. Great. Absolutely. So we don't need to set that. It's all done for me. Okay. Deploy me. Woah. That's a lot. Do you have enough, like, in your current time? It's actually one core. Right? We we'll see. We'll see. Yeah. It's one core. Yeah. Absolutely. We'll see. We'll see. No problem if it's not the case, actually. I mean, this is a this is a nice machine. Lucky you. I'm the CEO. Supposed to have course. Yeah. Now I'm the CEO. I'm not supposed to have, like, a cool machine.

38:20 Yeah. Although you can see, it's just all good Google Chrome. Yeah. I don't even use Chrome. I use Firefox, but I have to use Chrome for streaming because not Firefox just can't do multimedia. It's awful. Awful. Awful. Alright. It has a port six six six. Let's publicly expose it. Yeah. And Cool. Cool. Yeah. Yeah. You you can set that afterwards, actually, so you don't need to necessarily just, yeah, do anything. I'm just clicking on stuff now, to be honest. You're right. Just I'm just curious. So You're right. We can set the interface and see if the front end

39:01 team is is good. Alright. Let's see what happens. Yeah. That's the most interesting part. So now what you have done is that you have created a service. And, actually, inside this environment, you can create multiple services. So your database, your different stuff. Oh, there is an error. That's perfect. Perfect. Okay. Hit one. So we got the It says to look at my deployment. I mean, I thought I was looking at my deployment. But, actually, it's just yeah. That's the yeah. You're already on that. I'm just I'm looking at the fade, but we're all back into the version.

39:48 And Built to matter image. What will be the error, actually? So if you fade mirror image, that's quite interesting. I'm not sure about this. So we can we can dig we we can dig into that together, actually. I'm just gonna make sure. Yeah. If you can just do that. But we found the the image. So Must be the one. Really? Yeah. Docker emojis aren't allowed to use mixed case. Oh, that's I wonder if that's the problem. That's that's that's a that's a good so I will report back that to my team. So they they need to be very good. Yeah.

40:40 So now you get to add add a service. Right? So Yeah. You you you can't expect, like, developers just digging into that kind of error. So when it happens, it's kind of it's kind of messy for them. Alright. So it's now redeploying. Yeah. Absolutely. So you can look at the logs. It seems happier. Yeah. You can look at the yeah, the details. Yeah. Seems to be much, much, much happy happier. Yeah. Well done. I'm now pulling it here. You're now Slowly pulling it here. But yeah. Yeah. Oh, come on, Internet. You're fast open this. Yeah.

41:39 That's a that's that's that's the counterpart of of running that on your local machine. But at least you you see and you get the experience, like, it's Well, let's do some hellos while we wait. So hello, Mozz. Welcome back. Good to see you again. Thank you. Yep. You were here last time too. Thank you for joining us again. Thank you. And we have a question. Yeah. Y m o wants to know if we're going to cover self hosting Qovery. So I know you set the control plan as an open source. What does that mean for people that are

42:12 in that kind of self hosting land? Yeah. That's a very good question. Actually, we are working on an enterprise version, like, for next year. It's in our plan, like, just to provide the ability to also self host the control plane, but only in enterprise, like, offer just for, like, security compliance reasons, mostly bank, for instance. They need that. So, yeah, that's the options that we do. Mozz has got a question for you. So does Qovery support deployment to instances or virtual machines? For instance, can I manage Ansible deployments with Qovery? That's a very good, very good question. I

42:55 like that I like this question. Yeah. This is real real world question, and the response is yes. Actually, Qovery is we use Kubernetes as, like, an orchestration layer, And you can deploy Terraform script. You can deploy your own Terraform module. You can deploy your CloudFormation script, Handsible script. You can deploy anything. And I'm gonna show you, like, a much more complete, if you want, a pipeline where we deploy such things like different type of stuff. We even have customers actually that they are only deploying Lambda application. You imagine? So, actually, they use Kubernetes just as an

43:32 orchestral year with Qovery, and they deploy Lambda applications for ephemeral environment to create dynamic environments for testing purpose. That's quite interesting. Right? It's different, like, from the main use case of using Kubernetes, but Kubernetes is really good at just orchestrating stuff. So Yeah. I think that the Kubernetes model, if you're comfortable and happy to write controllers that can live in there and engage in that fashion, you can do a lot and build a lot of really powerful software on top of it. You know, I was just to talk about SaltStack that like, before Kubernetes, SaltStack was one

44:09 of my favorite pieces of software, but I excelled at that exact use case as well. Yeah. So Yeah. We have a deployment. That's very good. So you can click on for the overview, for instance. And this is something interesting. So now you can just look at the you can open the links. You see the links on the the right the top right? Uh-huh. Yeah. You can open this link. It will work. Is that not found? Like oh, maybe your machine is not like do you ex oh, yeah. Is that supposed to to work like these

44:41 things or not really? Yeah. Yeah. Okay. I mean Is that working on your can you just look at the yeah. Quick. Quick. Quick. Tick. No. I that's not gonna work because it's a cane cluster. That's my oh, that is the cluster. No. No. It's not the cluster. Yeah. You can see the application. So it's running here. Yes. Mhmm. That's good. And Okay. So let's do Yeah. Forward. Yeah. Let's do that, and let's see. Apart. 666. I'll just pick around 8080. Oh. Not found. Oh, the namespace? No. No. You you missed the maybe the six. You still need No. It's just the namespace.

45:41 Okay. But the part is, like, four times six. Right? That's I know. I know it's three times six or four times Three sixties. Because it's part of custard, so it's the number of the devil. Oh, okay. Yeah. Yeah. That's a good one. So that field for binding reasons, means I probably got that the wrong way around. Right? Because it should I can never remember. Okay. So local host. Yeah. Yeah. It it works that way. Can you try something for me? You see you have the access info. Just try something. Yeah. You can local machine. So you have if

46:36 you you have a tab that is local machine actually, so, Qovery, we do provide, like, a way just to pull forward through our CLI, and we have a system that makes makes it possible because we install a REST agent, like, that is forwarding the thing through the control plane. If you execute that command, maybe that should be the p, so it's something that is a bit dumb, but we at the local. And yeah. Exactly. Okay. And now if you are doing local parts Okay. That's that's probably I don't know what's maybe there is something that is not rooted,

47:26 like, properly, like, through this, yes, disengaged. So but, anyway actually, it works, so it's well it's well, like, exposed. Right? Yeah. And that's something you see. Actually, if I'm look so that's quite fun, David. If I'm doing this command, you give me access to your organization and you give me the permissions to do that, like, to your application. From my machine here, I can connect to your your application. Nice. So you're running some sort of proxy in the cluster that Yeah. Handles all of the traffic. Okay. Yeah. Yeah. And it's the reverse connection to the control plane. So everything

48:02 is transiting, like, the control plane, but nothing is stored in the control plane because we don't store any information for security purpose. So it's all just proxy. Okay. So I'm I'm I'm curious. Right? We have deployed something. Yeah. We've got a bunch of variables, some of which I set myself. Yeah. But I didn't actually see a way to expose these to the workload? To expose the the environment Yeah. How how do I add the environment variable to the workload? Actually, they are they are. Like, they are supposed to be in the workload. So Oh, they are? Okay. So that's just done

48:52 for me. Yeah. The yeah. Again, put yourself in the shoes of a developer. They add environment variables. They add secrets. They want just to see them. Alright. Then I put this to you then. One of the biggest frustration and I work with a lot of people that need help with Kubernetes. Yeah. One of the biggest frustrations is they have an environment variable or a config map, and they come in and they say, okay. Actually, that value is wrong, and they save it. Okay. So you offer to redeploy that for me? Yeah. Yeah. Absolutely. Yeah. Is there any status here that says

49:29 this is sync? Yeah. You have, like, you are supposed to have a probably, like, you just a bar, like, saying that you have to redeploy. For the environment viable, maybe not. But if you are changing any things like else, yes. But, yeah, you get the live status. So if you are redeploying, you get the live status and info from your your application. Yeah. Okay. So okay. Now it's telling me that this needs to be deployed for those changes. Yeah. Okay. Alright. Go for it. Yeah. Absolutely. Yeah. So there is something also, like, that that you you didn't click on is the service

50:10 logs. So you can see also, like, the logs in real time of your application. You have the tabs just to write to to deployment logs, actually. So, yeah, it was just next to Mhmm. You have two tabs, deployment logs and service logs. Where's the service box? Oh, yeah. It was, yeah, it was exactly in the same. You see, you have two tabs here. No. Actually, when you are just in that in that console screen, there's there's two tabs. I know that it's a bit confusing. That's something that we are working on, but you can just yeah. If you go you

50:52 go back, like, to the logs. Yeah. Oh, right. Yes. Yes. Exactly. Right. Exactly. That's a bit confusing. Yeah. Yeah. Yeah. Yeah. Yeah. I know. Yeah. I completely missed that. I just Yeah. Right over my head. Service logs is the logs, the live logs. Like, if you were, like, doing Kube CTLs logs, basically, that you get, but in much more, like, a detailed way. Yeah. Yeah. Absolutely. That's nice. Yeah. And this is transitioning from directly the so from the control plane again. You see? So Oh, I can switch pods. I mean, I've only got one. Okay. That's

51:37 a really cool feature. I wanna play with that. So I'm gonna come into my settings, and I'm gonna say, how do I scale this up? Give me three, five. So I'm assuming this see, the existence of InstanceMen and InstanceMax tells me that there's probably some sort of auto scaling that's being managed for me. I haven't seen a way to set thresholds or settings yet, so maybe you can lean into that a little bit. But I just wanna spend up more than one because if that does what I think it does from the terminal point of view, that's really cool.

52:10 But maybe you can tell us a little bit about where was that thing? Services, cluster Yep. Settings. Okay. So we have the ability to set meta and max. Do you handle auto scaling as part of your control plan? Yeah. Yeah. Absolutely. So the way it works is you can so you have the default values. You can't cover it as a video game. Okay? We cover, like, 90% of your need out of the box. Okay? We are just, like, doing the things the way that most people do. Then if you need to change some internal parameters, you can get into the

52:46 advanced settings. So you see the tab that you have on the left, advanced settings? You can get into that. And this is where you can start to tweak, like, the internals and change, like, some specific parameters. It's a key value stores like a system. And, basically, you have HPA. Like so if you are looking so if you search for HPA, you can you can find, like, the the the way just to to change also, like, the percent like, the CPU percentage for the auto scaling part. You see? So, yeah, those are elements that we add

53:26 depending on, like, what our customers want and yeah. Exactly. Recreate recreate is is available. I'm just gonna break stuff now because it was fun. No. That's good. That's good. Breaking is good. I've turned on read only file system. The application doesn't touch the file system, so we should be alright. I changed the deployment model to recreate, which should speed up my deploys. But, obviously, I'll I'll sacrifice and accept some downtime. I don't wanna redeploy it just because, you know, we have a deployment that I wanna play with first. Right? So we can now see that I have three pods.

54:02 Yeah. And the thing that piqued my interest was this. Yeah. That's so oh, now I'm in the wrong pod. Okay. Let's pop over here. Ah, that's nice. Okay. Yeah. Exactly. Okay. That's a nice it's such a I'm I'm sure it's a simple feature in the back end and all that. But to me, that's just one of those, you know, when I find a quality of life thing like that, it just makes me happy because it shows that people understand how hard it is to debug, and you need to be able to switch context, especially across pods within services.

54:37 Yeah. I'm so interested about it. So Even even developers. And just to to tell you, like, this feature is not, like, an easy feature because it works again, like, in real time. It's IO, and we are playing. So I know that my one of my staff, a back end engineer, explained how we he did that. So it's written in in Rust again, and he's interacting directly with the Kubernetes API just to stream the the input and the output. But you have, like, some restriction, and, also, you need to transfer back the information to the control plane. The control plane

55:10 needs to transfer back those information in real time via WebSocket to, like, the the browser. You see? So there is some challenges, and we have some customers that they are doing some crazy stuff, like just streaming, for instance, megabytes of data, you know, doing those tail stuff. And and this is the type of, like, experience that just overload, like, the whole system that can overload the system. So if it's not, like, tailored, like, the right way. But, anyway, that's I am now watching it stop all three instances at the same time and, hopefully, spin up three more for some pretty tightened up

55:50 resource restrictions. But let's tackle a couple of questions just now as well. So Mozz asked, is there an ingress controller here? How are we handling traffic to the cluster? Yeah. There is an ingress controller. Yeah. Absolutely. It's a ingress controller. And, actually, when you install Qovery on your Kubernetes cluster, you can tweak. So something that you didn't see, David, because you are using the demo mode, is that when you do generate, like, for real, the configuration with the CLI, the QCRI for your Kubernetes cluster, actually, Qovery do generate for you the final hem charts, like hem values file. So you

56:27 can apply, you can change the parameters, and you have access to those parameters like the engine's eGAS controller and so on and forth. Yeah. I wanna make sure I understood that. So Yeah. Let's pop over to production. Let's say new service. You're saying if I pick a Helm chart, which I'm assuming Postgres, is that a Helm chart? So it's a it's a yeah. Under the hood, it's a it's a hand chart. You're right. You're right. But here, it's a it's a prebuilt, like, hand chart. So we are doing so you see, like, this fancy interface for developers that

57:05 they can use. But if you want to deploy your Hemcharts, you if you want, you can go just up, and you see you have the Hem yeah. And Hem integration is very good. Like, it's very people, most of the time, when they see how we deintegrate that, they they really love, like, even DevOps engineers, actually. Yeah. Oh, I don't know where they are. So it's TTPS probably, I do think. Yeah. Art backup. Let's see if we can get our final one. Yeah. And we have their own chart. Okay. I love this product. I'm spending 20%

57:52 of my day on Grafana looking at statistics of stuff to report. Yeah. It's fantastic. So alright. So you got the repository. You scanned it and pulled down all the charts, which I'm appreciating. Nice. Yeah. It's Grafana. Yeah. Oh, let's go with Grafana. Latest. And something that you didn't notice, probably, David, is that when you are looking, you know, all those elements, version, Grafana, the chart name, you can type directly with your your keyboard. Actually, it will work, and it will filter automatically, like, the the elements for you. Ah. Yeah. So you can type directly the the

58:35 comment. That's something probably we need to improve in the design system that we have. Yeah. Oh, nice. Okay. So I break everything. I I Yeah. Oh, okay. So if I have a values file in either my own GAML or a public repository. Yeah. Absolutely. And you you want to see something very nice? Yeah. It's the whole mode, like the whole YAML mode. So he yes. And you create override. Yeah. And here, we do load to have the different values for you. Yeah. And you can create your own one, like, on the left. So it's good, like, just

59:20 to kick start quickly and do the stuff. You need to add a feature where there's a little arrow here, and I click it, and it copies it over here. That's that's a very good yeah. Absolutely. Yeah. Yeah. Definitely. Alright. I'm not I I want I think you might for the developer, they will just hate you. But at least in terms of experience, it will be very nice. Alright. Go. Alright. Another question from Miles while we wait for to to roll out. Do you support backup and restore of persistent volumes? So we have also, we support, like, persistent volume

1:00:08 first. So you can add persistent volumes to your application. You have an option to, like, all the settings of your apps. Like, yes, the settings, storage. Yes. Exactly. Oh, that's something that is disable disable, like, probably. Okay. Yeah. Probably because it's the demo mode, and we have, like, some features that you can add on the end values, because you need to define, like, the storage type. I do think that's the reason. In terms of automatic backup and restoration, no. We do not support that. We didn't have, like, any request, and that's the way also we work is that we

1:00:43 obviously, there's so many things that we can do. The way we work is that we have a public road map, and we are very close to our customers so they can just report back, like, the things that they want or they might want, and they can do that. But if there is a specific features into Qovery that makes, like, everything possible, is can you just show this maybe, David? So if you go back to create a service instead of creating a so, yeah, if you go back here, new service. Yeah. You have the life cycle job. The

1:01:16 life cycle job is a is a is a is is basically a container that you will provide so it could come from, like, a git repository or container history where we'll execute that task. So you can specify actions, and we can and you can build your own pipeline and your own stuff. So, basically, you can execute comments. So let's say you want to automatically backup and restore stuff and you want to create an action for that, you can do it, and you can automate the whole process. That's the way that, things like everything possible, basically. Yeah. I hope I'm responding to

1:01:50 your question, Moses. Alright. Thanks. Yeah. I didn't really when we came to this page, I didn't pay a lot of attention to the first ones. I went straight down to, like, okay. What's difficult to deploy? But, yeah, life cycle jobs are that's a neat feature. Databases. Yeah. Nice. Okay. Yeah. And when you are running on your own so if you use the query manage option, let's say you run that into AWS, you will get access to the manage databases as well. So meaning that here, we'll in just one click, you can put in manage database, manage

1:02:28 progress, RDS. Right. Okay. Yeah. Which is convenient, like, for companies. Again, remember, like, the perfect scenario is companies that they are coming out of past. They wants to move into The US. They are not super familiar with all of that, so they just wants to to deal with that. Yeah. Alright. So, again, I'm just clicking on stuff, but I'm curious because Tell you. You're curious. Well, I mean, that just means I'm excited. Right? I'm enjoying what I'm seeing. So having built a lot of platforms, what happens is, you know, we roll out production. We're all happy with it. At some point,

1:03:10 we need the staging environment. And what I was hoping to see when I clicked I mean, export is TerraFarm is nice. Right? I wanna take into that a minute. I was hoping for, like, a clone to staging. Luke, you have you have the option. Oh. Oh, you just made me happy. And I can change the cluster so I could close that to AWS or somewhere Yeah. Oh, look at that. That is a that is a pure joy smile. Yeah. That's a very, very, very we have even Nasdaq companies that are using Qovery just for that feature, just to tell

1:03:48 you, Because they can create their own environments, and they can duplicate for their QA, sales engineers, and stuff like that. They can just duplicate stuff once it's defined. Okay. And let's click on export to Terraform now because I'm curious what this does. Yeah. Absolutely. Downloads. Unzip. Where did it go? Oh, no. Send this to alright. Okay. Main variables. Sorry for the light mode, everyone. Yeah. Okay. So that is Qovery automation for what I've been doing on the UI. Right? Yeah. We have a telephone provider. That's one more reason why DevOps people, they like it because they can

1:04:43 control it everything through our telephone provider. That's usually for pollution. That's what happened is that at some point, they do they start fast, but then they move into, like, a GitHub's way of managing everything. And that's very convenient here because they can start at least from something and then adapt to what they need. Okay. So this is a Qovery container. Yeah. Yeah. We do maintain that for our customer. Yeah. It's it's our own provider. Okay. So now I mean, I'm I'm more curious now. Production, clustered, settings, storage, no. What happens if I then add a volume

1:05:30 to the TerraForm HCL? Yeah. Will it work? If you add, like, a storage now, like, an object, like, into that? Probably not. I do think you will be rejected by the by the API, actually. Right. Okay. Okay. I was just curious. Yeah. I mean, it's nice that we can eject away from the UI. And, again, you know, the best can live and get Yeah. And could be cloned and and reapplied across different clusters and stuff like that. But then I think when the UI is right there, it'd probably be inclined just to kind of stick with this. I mean, the export is

1:06:10 nice. I I could have snapshots of point in times, but I I I wouldn't I'm not sure I would wanna jump back down to modifying HCL. Yeah. That's I guess that's the choice. Right? People can decide what they want and what's best for them. Yeah. Absolutely. And something that we are working on is that we are more and more we have more and more, like, software engine, one side, DevOps that they're using the platform and controlling it. What we want to do is that when a mode so that's something that will come, like, next year. But, basically, when you modify the web

1:06:39 interface, instead of applying the changes Yeah. Instead of applying the changes right away into the API and doing those changes, we'll to deny, like, the top configuration and update, like, the gateway, like, appearance of it in the overview and then apply. And then Yeah. Usually, people today, they choose one or the other. Nice. Alright. I think we've kicked the tires on we I mean, we don't even use the get started page. Right? We just let me go click happy on your UI, which is which is nice. But I feel like, you know, we have showed people a small

1:07:21 contrived example, but powerful example of what people could can do with Qovery. Is there anything else that you want to show people that gives them a snapshot or a look into what you or your customers or production deployments look like with Qovery? Yeah. I can I can show I can show maybe, like, a few things? So maybe a much more like a so maybe a much more detailed configuration. K. You happy for me to to throw your your screen up? Yeah. My screen yeah. Yeah. You can. Yeah. Absolutely. Yeah. Yeah. I can I can show, yeah, I

1:08:02 can show a much more, like, bigger view, like, of different type of services that they're running on on on the coverage? So this is a an environment containing, like, a production, but, actually, it's not. It's just for my my test. But here, you have a bunch of so it's running on AWS on a managed cluster. And, actually, this cluster has even, like, Kubernetes Carpenter turned on. We have an integrations with Carpenter, so we do install and manage that with our manager for. So meaning that I can benefit from the spot instances and the the downscale part, like, of this, know, the

1:08:36 the scaler, which is great. K? And, basically, here, you have, like, back end, front end databases, even Terraform script. That's what I wanted just to show you. So you see I've deployed RDS instance with my own Terraform module, and this module is just is just that. You know? Look. It's a pure Terraform script. Okay? So where I declare my resources, and I have, like, some inputs and some, like, output and variables. So I can define that. And you see, the beauty here is this. Because once you have executed this Terraform script from Qovery, you want to get back the informations from, like,

1:09:17 the endpoint, username, password that they are do generate it at the end time. And even that, we do manage that properly into Qovery. So when you start deploying Terraform script, look. I got the job output values that they are available, like, at the execution. And that's what makes, like, Qovery so powerful because here, actually, what you get is a deployment pipeline. So all my services, I can just execute in the right order. I can deploy in the right order of my different services, so even my Terraform RDS. And when it happens, like, Terraform is that

1:09:53 those output values, they are injected into the other services. They are available to the other services at the time of the execution. So you see? So that that's something that is very, like, much more much closer than than ever, like, to the to the production to something that runs, like, from our customers. And, yeah, I just wanted to show you that because, obviously, when you you deploy one application, it's very simple. But when you need to interconnect, like, microservices applications altogether and just order the right way those different services, then this is why it's, like, much more harder.

1:10:31 And we cover, like, all of that, obviously. And you can clone just for your own you can clone even that environment, David. So I can choose, like, even another, like, a cluster. So for instance, let's say Tokyo cluster on Google, I can just clone that one. Obviously, the RDS, so will not be, like, properly working. So you you can adjust and adapt, like, based on the the meta informations that you have from Qovery. But, yeah, you can clone, like, all the resources. Very cool. I like that. Yeah. And if you just go to that pipeline

1:11:10 where we see that phased rollout Yeah. If you go to the pipeline, you get exactly the same. Alright. So it like, will that update and say, oh, there's an SQS there, which isn't gonna work on GCP. Right? But, you know, I was curious if it visualized the progress through the pipeline from left to right. No. No. No. But that's something that we want to bring, like, a much more, like, visual view of when something is deployed. But today, when you do deploy, a a pipeline so if I redeploy this, for instance, pollution, you you you can see, like, from the

1:11:48 logs interface, you have a view Nice. Of the pipeline. Yes. You have the first stage here that you can create. You can customize everything, but you have the first stage, then the second stage, third, and so on and so forth. So you can see you see, boom, here, like, the the deployment of the different resources that they are done in parallel in each stage. So if something goes wrong and that's something that is very nice because when you deploy an environment, like, your service can go wrong, so you can just see exactly which one at which time, and you

1:12:20 can troubleshoot that service, and you can redeploy just that service or set of service and so on and so forth. Awesome. I like that. Alright. I'm gonna push this back to is there anything else you want to show before I I push this back to back face mode? No. Actually, you know, otherwise, we can spend, day. Yeah. I mean, it was it was daytime when I started the session and it's Yeah. That's good to hear. Alright. Let's wrap this up. Maybe you could share a little bit about, you know, I I think you've you've given us the history about

1:12:55 you and and Qovery. We've seen the product. I think a lot of people are really gonna enjoy what they see. What's next? What what what's what's on your road map? Yeah. Yeah. That's a very good question. So the goal for us is to use, like, the integration as much as we can. We have worked on the AI agent that is open source and have, like, our customers to transition very easily, like, from past solutions to, yes, using covering in just, like, no time. Today, it works from to AWS and the different those account provider. In just five

1:13:26 weeks, we can move out the workload. That's crazy. That's something that were not even possible, like, before LLM. That's really it's interesting. We can drop the link because it's available on GitHub. But, basically, we focus on that, And then there is, like, many more that we we want to provide, like, supporting also providing, like, Azure managed, like, offer. That's something that we are working on. And you talk about that, but those enterprise features that we want, like the self managed control plane, that's something also we want to deliver. But, yeah, the product is, like, evolving all

1:14:02 the time. We have invested a lot, like, in the developer experience, and we are eager just to get feedback from users when you see something so we can just, like, work on it. But those are, like, the main the main big things that we are working on. Yeah. Nice. Awesome. Alright. Well, I mean, thank you so much for taking time out of your day to sit here, talk about your product, show it to people. And, yeah, it's been an absolute pleasure. I think it's a fantastic developer experience and user interface on what is a pretty cumbersome and and difficult challenge for

1:14:34 many people and many teams and many organizations. And I hope that it will vary for the people that are out there going I need this. We'll solve some real challenges for them. So, you know, thank you again. Yeah. Thanks, David. Thanks for letting me take a stab at everyone. Alright. And then thank you to everyone for watching, you know. Feel free to say goodbye in the comments as we write this out. We'll see you all next time. And, Rawkode, thank you again. I'll speak to you soon. Bye, all. Goodbye. End of stream, let's leave our Rawkode

1:15:11 Academy session done. Now we wait for the next one. Learning cops, we all connect. Knowledge gained, we all reflect. Tech talks and the coding grind in the stream we all didn't find. Thanks for joining Vans for the thrill. Every moment, every skill. Hope to see you next time, Rawkode Academy. Tech about