Kubernetes

“Standard 19 inch rack, standard air cooling, no complicated liquid cooling retrofit in the data center. We're using standard Kubernetes, standard Red Hat Linux, standard Ethernet at the top for networking. We don't have to use all this kind of really expensive…”

“Kubernetes is hard to manage. It's not built for burstiness and custom images and has a terrible developer experience.”

“We think that the same benefits that apply for DX also actually apply for AX, which is why would you have an agent read through hundreds of Kubernetes files and, like, write YAML that's not even typed when it can basically make a couple of changes in a decorat…”

“And for render, that was never an option. We said, look, if you have to use Kubernetes for something, that's a failure.”

“In my past life I worked at Heroku and like Kubernetes was much more successful than Heroku. Even though I think from a user experience perspective, it was much worse, but the Delta was Heroku was saying, Hey, we're going to replace your ops team. And Kubernet…”

“We had Kubernetes in the cloud, but Kubernetes is not how you deploy your AI imprints and your AI training workloads.”

“Dynamo has a set of components that A, tell you how to scale. It tells you how many pre-fill workers and decoded workers it thinks you should have. And also provides a scheduling API for Kubernetes that allows you to actually represent and affect this scheduli…”

“I think there's a somewhat similar situation here with the caveat of saying, like, 10 times faster. Like, just day one, so much momentum around MCP. So much interest in this. And then also, 10 years of CNCF to sort of teach the developer community and the vend…”

“It was less number and also it was mostly lift and shift. It was not container, it was not cloud native applications. Cloud native applications started happening mostly around 2018 when Kubernetes came.”

“Google historically, when it came to AWS, Open source Kubernetes and went after them. When it came to Apple, they open sourced Android and certainly some of their lower models are open sourced and competitive on open router, but maybe they should be more aggre…”

“Inside Google, they still run Borg. They run far less Kubernetes than they run Borg, and Borg is part of the secret sauce.”

“Now, the interesting thing is the industry, nobody made money on either Docker or Kubernetes. I mean, there were some startups, right? But really, there was no big company that was created out of that.”

“We're seeing more and more enterprises that are actually comfortable with just building on top of a simple Kubernetes abstraction or something. And basically, you know, cherry pick a couple of snowflake or database type services on the side that help them comp…”

“I think a good analogy here is sort of, like, technologically, it's kind of, you want to be the Kubernetes of the world for the agent, but with much better DX and easier, easier to use.”

“Google had a Piece of technology called Kubernetes. It was orchestration that would allow you to move a workload. If that became a standard from one, you know, one large server vendor to another, right? It basically created ease of distribution. So you could r…”

“The thing is you only spend about 10% of your time doing that, you know, as a software engineer in practice, you probably spend about 90% of your time, you know, dealing with your Kubernetes and fixing your unit tests and upgrading your thing to the new versio…”

“See, bro, the first step was making sure you don't have to use Kubernetes, and that we kind of solved. Now we'll do the rest. Like, that's just it. Like you can run agents without Kubernetes or VMs.”

“It's easy to conflate long running with CPU intensive, and these are kind of not the same thing, and actually a lot of Specifically agentic workflows. A lot of what you're doing is you're waiting. You're waiting on human in the loop. You're waiting on LLM. You…”

“And then you could speculate, can the agents maintain it? But I think. I think that, that then we are definitely in a 15, 20 year horizon before, before they could be relied on for anything like that.”

“We can't really use Docker and Kubernetes. So we're going to have to throw that out and probably going to have to build our own file system, which we did pretty early in a build our own scheduler and build our own container runtime. So that was like basically …”

“I think it is one of the most savvy competitive Weapons you can bring to the field. I'm sure they were worried that open AI would create proprietary data connectors and they want everyone to use this one very similar to what Google did with Kubernetes.”

“So we built our own, we threw out Kubernetes out the window. We threw out Docker out the window. We built our own File system in order to optimize for how container images are distributed. We built our own scheduler in order to maintain this pool of workers an…”

“Less layers of infrastructure, less layers of abstraction, make it a lot easier to work with. Like we don't use Kubernetes, for example, I would just directly launch these things and it's just been much easier to debug this way.”

“And they did some of the work that they did in the size of that team. They did the equivalent of Kubernetes and what Google was doing with buildings full of people with 10 people in a room. Still, I don't appreciate, I don't know that Technical brilliance of t…”

“Like in particular, like Kubernetes, I feel like it's like kind of worked okay for backend teams, but not so well for data teams.”

“Every MLOps team then realizes they could raise like a shitload of funding, you know? And so like you got like every, every major company, their MLOps team like went off and like raised money to like make a new product in the market, which I think from an inve…”

“So the developer experience got way more complicated, right? So like, you know, we have just like a TypeScript engineer who does front end, and suddenly just to set up his development environment you know, he's gotta understand how Kubernetes works, or, you kn…”

“And this was started around a year ago and, you know, from zero, they went to 17,000, I think now 18,000 GitHub stars, which is a lot, right? Like, I think, if I'm not wrong, Kubernetes total has some 30,000 GitHub stars. So, like, that shows you, like, how, h…”

“Kubernetes clearly which was built by Google, is the standard now for doing this”

“Knowing a little bit of like Go or like having like spun up Kubernetes on my, you know, on a test, test cluster at home is not nearly enough to have an informed good opinion about A new language or like a new system. Like you have to run those systems in produ…”

“Implementing those technologies does not give you those outcomes we talked about. We actually looked at people doing mainframe stuff. You can achieve these results with mainframes. Equally, you can use the, you know, Kubernetes and, you know, Docker and micros…”

“Google just gave Kubernetes over to the Cloud Native Computing Foundation. And that's like, you know, an up and coming project. It's getting lots of contributors and lots of new adoption. And that was just because, like, for it to be a de facto standard, it ne…”

“I firmly believe that, you know, open source has become the, I think, the dominant way to distribute, you know, software infrastructure, you know, whether it's you know, something like Kubernetes, or it's a database.”

“Over time, we will be building out a set of enterprise grade extensions to the platform that will be licensed product.”

“And so Kubernetes has really, you know, provided the step from having to think about virtual infrastructure to thinking about this logical infrastructure, this perfect idea of what a machine should be. And that's largely how, you know, developers at Google and…”

“But I would say anecdotally, at least, ah, two thirds of the Fortune 2000 are using Kubernetes in a material way right now.”