 What we're doing is, we don't tell you use our product, it is your platform, we don't do that. What we essentially say is, here's a building block that would be helpful for your platform. Vcluster, Devpod, those are great building blocks for a platform team to build on top of and essentially compliment their platform with these capabilities. Hi, this is your host, Abdul Bhartia and we are here at KubeCon Chicago. And today we have with us once again, Lucas Gently, CEO and co-founder of Loft. Lucas is good to have you back on the show. Yeah, good to finally have a session with you in person. I've only seen you virtually so far. I mean, we've met at KubeCon before, but never for a recording. And interestingly that the company was created, we have been talking since the very early days and also we have seen the growth of the company. Now you have a great booth here, Gold, right? Oh yeah, we upgraded to the Gold sponsorship. Exactly, so I do want to talk a bit about the company before we talk about Vcluster Pro and technology spec. Let's talk about the growth of the company that you have seen in these years and what has been driving this growth? Two and a half years ago, we were just two or three people to start out with. Now we're over 25. We've seen a massive amount of growth in our open source project, Vcluster. So much adoption. We've seen commercial success with some of the largest enterprises in the world. And it's just fantastic to see. And I think there are a lot of KubeCon talks that mention Vcluster. There may not be specifically about Vcluster. There may be a result about something else, but then people mentioned that they're building this on top of Vcluster, which is even more exciting to actually see. And which actually brings us to Vcluster, Vcluster Pro, you folks and all. So talk about the evolution graph of Vcluster itself. Open source all the way to the commercial offering because we discussed earlier also, open source has its own limitations, because customers watch more. And because of the nature of open source, you can do only so much because then a lot of other folks don't want it. That's where the commercial. So talk about the evolution of Vcluster Pro. We started Vcluster two and a half years ago. And within the first 12 months, about a million virtual clusters have been created. And we thought that was amazing. That was a really great milestone. Just a couple of weeks ago, we hit 40 million virtual clusters. So now two and a half years in, you can see the growth trajectory in this open source project. It's crazy. I feel like we're in a similar situation as Docker was in the early day or probably back then VMware was in the early days. Initially it's a lot about convincing people that this is the right architectural choice. And that makes sense to add yet another layer of virtualization, right? Initially people are like, do we need another layer? It sounds complicated. Why are we doing this? Who manages this? But then they see the benefits of it and they give it a try and they see, whoa, this is actually going to be the way forward. I think one interesting thought I had throughout this conference while talking to people is if you're in an enterprise today and you need a server, nobody is going to go into a data setting anymore and plugging in a server for you. They're going to spin up a virtual machine. And the same thing will happen for Kubernetes. In five years from now, if you need a Kubernetes cluster, you're going to get a virtual one by default. And then only for the edge cases, someone goes in and plugs a server for you, someone really gives you a real Kubernetes cluster because the virtual ones are just so much more cost-effective and so much easier to spin up and maintain long-term. And then you mentioned, we launched Vcluster Pro, really, really exciting step to bring Vcluster to the enterprise, give them the confidence to run Vcluster in production, massive amounts of workloads. Really exciting what we shipped with Vcluster Pro a couple of weeks ago. I remember when you talked about, there was a talk about Vcluster and somebody from AWS was interested in a lot of, so community is driving a lot of things. So I also want to understand how community has received it. Yeah, the community is just fantastic. It's, we have like over 2,500 people in our Slack channel and you see just the interactions there. People help each other, people create, pull requests and they open issues and suggest new features. Folks are just really, really excited about this project. Everybody really wants to grab these hoodies and t-shirts that we have. They identify with the project, they love the technology and they are part of it. They were early adopters, they're really excited about it. And I think it's fascinating to see that so many people believe that Vcluster is the way forward. When we look at Loft, when we look at Vcluster, you have a lot of other open-source projects. Talk a bit about the changing market. We can talk about platform engineering, we can talk about DevOps. Where does all this offering fit? We can look at it from a myopic view of Vcluster Pro or we can look at it a broader frame of Loft and its role in the ecosystem. Obviously, platform engineering is the buzzword this year at KubeCon, right? We always have a buzzword. It may be DevSecOps or back in the days, it was more DevOps, right? Then it was multi-tenancy for you, then developer experience and now we're platform engineering, that's okay, right? I think the big difference of what a lot of people do is they phrase their existing products in the term platform engineering now. What we're doing is we don't tell you use our product, it is your platform, we don't do that. What we essentially say is here's a building block that would be helpful for your platform. Vcluster, DevPod, those are great building blocks for a platform team to build on top of and essentially compliment their platform with these capabilities. But we never say this is the platform, this opinionated way of how you should be doing things, which I think is very different from a lot of messaging that people are driving at KubeCon these days. Since we are talking about labels, jargons, new buzzwords, from cultural change perspective, what kind of changes you are seeing because you have been part of this community ecosystem and you're also playing a very critical role there. So talk about the cultural side. Yeah, I'm super ecstatic about the whole CNCF ecosystem and you come to KubeCon and I feel like Amsterdam was amazing, this one is even topping it. I think we're fully back from the COVID recovery, right? Detroit felt still a little half baked, but the past KubeCon and Amsterdam and this one in Chicago is just phenomenal. And it's really crazy to see that a community has formed around Kubernetes and the tooling around cloud native open source solutions and related vendor products that always comes together and there's so much inspiration and knowledge exchange. To me, KubeCon is always very, very exciting, but also very exhausting because you have like a million different ideas and conversations, but yeah, it's something I always look forward to twice a year. And I want to not talk about the different kind of culture which is the culture within companies. We talked about platform engineering. What kind of cultural change you are seeing because as you already said, there was a theme about DevOps, DevSecOps and every year it's less to do with a new label. Sometimes it's more to do with how companies are leveraging. So what are you seeing there? Yeah, we're definitely seeing that it's not blindly shift left anymore. I think that was, it's another buzzword from, you know, a few years ago, I guess, everything needed to be shift left, right? Security, provisioning, admin responsibilities and then people realized, oh, maybe that slows our engineers down and maybe not everything should be completely shifted left. And I think today we're shifting the right kind of things left and other things we're trying to standardize on. I think that's really what platform engineering is. In my opinion, it's on the one hand side, you standardizing things, you're streamlining processes, but at the same time, you're trying to retain the autonomy and the velocity that your engineering teams need. And that's a fine kind of line to walk. That's something you've got to balance. And I think the trend towards platform engineering in a lot of these companies is emphasizing the importance of, hey, we got to standardize and we got to put our compliance and security and provisioning standards in place across the company, but without actually hurting developer experience, without slowing anybody down. I think that's the really big benefit in terms of mind shift that has happened through platform engineering. And we're early. Like there's so many companies that are just forming these platform teams and they have questions like, does a platform team have a separate budget? Or is this IT or engineering budget? Where does it come from, right? It's still very, very early days. We see that with tons of our customers just starting out with the platform engineering journey, but it's an exciting one to start. Do envision a new term, New Jordan? I'm sure we'll have a new term next year. I also want to talk about Corvive. You folks partnered with them. You also had a session here with them. Talk a bit about this partnership. What it is about and how it's benefiting Loft and Corvive. Yeah, Corvive is a really exciting company. The first time we met them was when I think they just had raised their Series A, they were a lot smaller than they are today, a lot less known than they are today. And when they told me back then that they're effectively planning the compete with AWS on GPUs, I thought that was crazy. I thought that was a really, really hard challenge. And I'm like, who wants to fight the giant, right? They do, and they're very successful doing it. And that's super exciting to see. I mean, now two years later, seeing them as a customer and using VCluster Pro as part of the Corvive cloud architecture is super exciting. The joint talk that we had, I think there were so many questions afterwards, right? People came up to us. I feel like every second person coming to the booth now has questions about like, how does that work? I want to dig a little deeper. I'm super curious. It's very fruitful, I think for us as a company. And for Corvive as well, I think, you know, one thing that gets me really excited about Corvive is they're very open. They build on open source and they're willing to share what they're building internally. They're going to KubeCon, giving a talk and actually explaining the inner workings of their cloud platform. Not every cloud vendor would do that, they do. And that really shows their commitment to the community, to open source, right? And to giving back, they've been amazing partners for us. And I think they mentioned that during their talk as well. They're planning to launch an open source project on their own. And I think that's really exciting. I can't wait for this partnership to grow even further. Just for the audience, what do they do? Corvive is a specialized cloud provider. They essentially let you spin up GPUs, mainly to run your inference workloads, your AI and machine learning applications. And, you know, they're building a lot of their cloud platform entirely on Kubernetes. I think that makes them very special. I think none of the existing legacy cloud providers out there can say that they're from top to bottom through the entire stack based on Kubernetes. Corvive is. And that's really exciting for us to partner with them as a really true cloud native, you know, cloud provider. We're so excited to be part of their journey. And how are they leveraging Vcluster? In our talk, we outlined this and I recommend for anybody to watch the recording, what essentially happens is their customers need Kubernetes. You know, a lot of these like leading edge AI companies like OpenAI, you look at the engineering block, OpenAI does pretty much everything with Kubernetes, right? So a lot of people that want to build AI workloads are obviously also wanting to build on Kubernetes. And Corvive hands out Kubernetes to their customers. And then the question is, do they create like a real heavyweight cluster for every customer? Or do they create lightweight virtual clusters with the capability to have shared nodes to put workloads on shared worker clusters, but also dedicated GPU nodes? And that's really important because if you're an early stage startup, you just raised your series A, right? You may not be able to afford dedicated heavyweight, large scale GPU nodes with a cloud provider like Corvive. So you may opt for the shared route and get that compute at a much cheaper rate. But then when you really need it and you want these dedicated nodes, you can add them. And the experience is exactly the same with Vcluster. That's why they're using Vcluster under the hood. I cannot not talk about the hottest topic which is generating AI. So I want to hear from you is when we look at generating AI, either from loft or from Kubernetes perspective, what does it mean? It could be two ways. One is loft or Kubernetes for generating AI or generating AI for loft. How do you look at it? Just like with pretty much any AI or microservice workloads, where to compute an architectural layer below, right? The fundamental piece that lets you run and scale workloads across servers and machines, right? And a lot of these AI and ML systems, they're very large scale distributed systems inherently, right? And that's what Kubernetes is really good at running these kind of systems. Also getting increasingly really good at running large scale batch workloads. There's a lot of emphasis on that stuff as well. And I think the other way around, I've definitely seen some exciting startups as well to build AI into their products to kind of facilitate managing debugging Kubernetes, right? But I think that's like, that's day zero. There's so much to explore still in that domain. We're not too deep into that space ourselves, but I've seen some really exciting booths here at KubeCon just walking by and chatting with a couple of folks. I think there's some interesting stuff coming in that direction. Of course, you folks have a lot of things in the pipeline. Some you can share, some you can not share because you're working on it, but just give us a teaser of what next to expect from Loft. Yeah, we're working on some really, really exciting features. A lot of integrations. Backstage, for example, is really on the top of our list because it seems very, very heavily used in the enterprise. And that's the plain field that we're in. And there's other exciting stuff coming directly into the clusters itself. One is the capability to snapshot and restore virtual clusters, which will actually allow you to transfer virtual clusters even between Kubernetes clusters or even between clouds. That's something our customers are really, really excited about. And yeah, I think we're looking at a commercial cluster API integration. So far, we have a cluster API provider in the open source, but no commercial support for it. And it's definitely not... It's definitely more of a side project at this point, but we want to integrate it directly into VCluster Pro and make it an essential part for folks that are betting on cluster API. So there's some really exciting topics that we're working on in that direction. Lucas, once again, thank you so much for sitting down with me and talk about all this exciting news. Curvy was an exciting one. And as usual, I would love to chat with you again. Thank you. Yeah, absolutely. Thank you so much for having me today.