00:13:11 · Anurag Goel, Jonathan Brebner
[00:13:11] S1: Yeah. Yeah. So the way I thought I would be able to figure out what I would feel excited about for an extended period would be to try to build things in specific domains. I decided that I I wanted to solve problems by building software. And I spent about six months learning all I could about healthcare. Spoke to a lot of people, including a lot of folks at the Commons, and then realized that the kind of software I want to build and healthcare problems don't go well together. At least at the time, that's what it felt like to me. And then I moved on from that and then looked at other problem areas like infrastructure for real time communications or deep learning. I actually built a tiny company while at the Commons that made it possible for people to spin up a Jupyter notebook. [00:14:05] S0: Is this the one that is between Render and Stripe on your LinkedIn? Yes. That... I was just looking at that earlier. I was like,
00:14:11 · Anurag Goel
[00:14:11] S1: what's that one? That is... That's... That was a thing that I decided to build because I got into deep learning, and I was taking this course online. And as part of that course, everyone had to spin up a Jupyter Notebook on an AWS instance with a GPU attached to it. And that whole process of setting up this notebook took people who weren't familiar with AWS. It took them sometimes up to a month, and it just didn't make sense to me. And I was like, this is stupid, so why don't I just automate this thing? And so I think I've created what eventually became one of the first Jupyter notebooks backed by a GPU in the cloud. And and then I realized... And it it got up to it got up to, like, you know, 10,000 data scientists using it just sort of organically just because it was so easy. This is where it goes back to what am I gonna be excited about solving for the rest of my life. And I I think as much as I love data scientists, we have them at Render. Building for data scientists was not really the thing that I found myself being incredibly excited about. So I went back to, well, what do I actually enjoy? And it turns out, as I had built all these things, I'd actually enjoyed automating the infrastructure behind building these products on the web. From my Stripe days even before that, I'd always cared very deeply about developer productivity. And so when you combine developer productivity and cloud infrastructure, that that combination seemed to really jump out at me, and that's how Render came about.
00:15:48 · Jonathan Brebner, Anurag Goel
[00:15:48] S0: Before we talk about that, I... I'm interested in the way you think about now and and then also at the time about the idea of sort of, like, killing an idea and moving on. Right? So the way you just put it is you just couldn't find yourself getting super excited about, about building for this particular kind of customer set or or... Right? But at the same time, you had all of these people who were... Wanted to use it, right, and were excited about it. And and that's... I imagine, like, that's difficult. Right? That's often what Yeah. What we have to, like, help people here figure out is is, like, I have some signal, but it... It's not quite enough that I'm, like, really convicted in the idea. How did you think about it at the time? Did you sit down and just go like, do I hate this or not, or or do I love this, or, like, was it that simple? [00:16:30] S1: Yeah. I think it was as simple as, like, is this the thing that I wanna be spending my life on? [00:16:35] S0: What was the point that forced you to ask that question? [00:16:38] S1: When it got some traction. Yeah. Because it's... Before it gets traction, you're just like, well, let me try to get some traction, see if it has legs. Yep. And then when it gets traction and you kind of see what it can become. [00:16:50] S0: You can project it into Because the future
00:16:50 · Jonathan Brebner, Anurag Goel
[00:16:50] S0: You can project it into Because the future [00:16:53] S1: until you have traction, until you have some amount of product market, you don't really know what's going become. And so after we had all these people using it, not we, I, I guess, either all we. That's when it became real. Like, okay, do I want to be growing this into something? And then I had this other option, which is, well, now I know how to build and I've built all these amazing developer facing convenient infrastructure products. And I've learned everything about all the backend technologies that were hot at the time. I realized that you can build something much more generalizable, something that applies to all software as opposed to just software for deep learning or Jupyter Notebooks. And that to me seemed like a much bigger problem, a much bigger scope, and something that could effectively keep me occupied for decades. [00:17:52] S0: We... We've talked about the kind of minus one is... And that's the way we've formalized the terminology. Right? Like, basically, what you were doing at SPC, helping to build the culture here, and then also figuring out what you wanna do with Render. We often talk about it as like, oh, it's what happens before you start a company. Right? And then you figure out what you wanna build, and then you... You're off to the But one of ways we've been thinking about it recently, and one of our other team members, Finn, wrote about this recently, is that it's actually a process that happens over and over again. [00:18:19] S1: Of course.
00:18:19 · Anurag Goel, Jonathan Brebner
[00:18:19] S1: Of course. [00:18:20] S0: Yeah. Right? Because you're constantly having to come up with a new strategy, new product, serve a new... Like, go after a new opportunity, and you're of, back to minus one again. What do we do? You know? Yeah. We have to build a conviction in some sort of way forward. Now seems like a time where that's probably actually happening more and more often. [00:18:36] S1: It's more true than [00:18:37] S0: ever. Right? [00:18:37] S1: Like, you hear about all these refounding of companies that's been around forever. [00:18:42] S0: Yeah. And just like the pace of technology Exactly. Picks up. Everyone's like like, okay. What is now possible with the latest version of, you know, this foundation model suddenly, like, all these things are now... Right? So how do you think about that, like as you have been growing the company? Has your approach to it changed? Are there specific things that you've looked back on and like, Oh, actually that thing that I did really early on is really useful as a way of thinking about doing it now? [00:19:09] S1: Yeah. Yeah. So, it's something that I think about all the time. It's something that our executive team thinks about all the time. And it goes back to, okay, well, every milestone you hit for revenue, it's 1,000,000, 50,000,000, 100,000,000, the next doubling, the next tripling, it's not going to be as easy because it's just the standard law of large numbers. [00:19:34] S0: So
00:19:34 · Jonathan Brebner, Anurag Goel
[00:19:34] S0: So [00:19:36] S1: how are you going to think about things differently? And that's where I think the minus one mentality makes a lot of sense because what got you there is not gonna get you to the next stage. This is true for startups just across the board and even for public companies, just any business because of competition, because of how markets change. And so this is why it's so important to be thinking about competitive strategy in terms of how you're different and where you want to be different. And for us at Render, it starts from ultimately who we're trying to serve. And you could change that, but then you're a very different company. I think Render has always wanted to serve application developers, people who are building these applications, not the people who are building AWS infrastructure, not the platform engineers, although they also benefit from Render. But at the end of the day, it's it's the folks who are building applications and these days AI apps and agents. And when we think about them, it's all a question of how are their needs changing. And we realized recently, for example, that of our top 30 customers by revenue, more than half are building some sort of LLM functionality into their app or building It's a [00:20:56] S0: thing nowadays. Yeah. [00:20:57] S1: Yeah. It's a thing nowadays. And this wasn't true even six months ago, by the way. [00:21:02] S0: Really? It's kind