0 to over $8M ARR in 2 months as a Claude Wrapper (Bolt.new, Qodo)

Hey everyone, welcome to the Latent Space Podcast. This is Alessio, partner and CTO at Decibel Partners, and I'm joined by my co-host Swyx, founder of Smol AI.

Hey, and today we're still in our sort of makeshift in-between studio, but we're very delighted to have, uh, a former returning guest host, uh, Itamar. Welcome back.

Great to be here after a year or more.

Yeah, a year and a half. Um, you're one of our earliest guests on, on Agents. Now you're CEO, co-founder of Qodo-

Right

... uh, which has been-- it just been renamed. You also raised, uh, thirty million dollar Series A?

Series A was forty.

Forty, forty million dollar Series A, uh, and we can get caught up on, on everything. Uh, but we're also delighted to have our new guest. Eric, welcome.

Thank you. Excited to be here.

Should I say Bolt or Stackblitz? Like, is it- ... is it like its own company now or?

Yeah. I mean, Bolt's definitely, uh, bolt.new. That's the, that's the thing that we're probably the most known for, I imagine at this point. Um-

Which is ridiculous to say because you were working at Stackblitz for so long.

For so long. Yeah, I mean, it was-- within, within a week, we ha-- you know, we were doing like, you know, double the amount of traffic this da-- and Stackblitz had been online for seven years, and we're like, "What?"

But anyways, yeah, so we're, uh, Stackblitz, uh, the company behind bolt.new. If you've heard of bolt.new, that's, that's our stuff, so.

Yeah. Yep. Excellent. Um-

I think by the way that the founder mode, you need to know to capture opportunities. So like kudos on doing that, right? Like you're working on some technology and then suddenly you can exploit that to new world. Yeah.

Totally. Yeah. I mean, you know, I think, uh-- well, not to, not to jump but-

Go ahead, please

... hundred percent. I mean, that's-- I mean, kind of a couple of months ago, you-- we had the idea for Bolt earlier this year. You know, but we haven't really shared this too much publicly, but like, you know, we actually had tried to build it with some of the, you know, state-of-the-art models back in January, you know, February.

You can kind of imagine which. And they just like weren't good enough, uh, to actually do the code generation where, where the code was accurate and, you know, as fast and whatever have you, without a ton of like rag.

But then there was like issues with that. So we put it on the shelf, and then we got kind of a sneak peek of, you know, some of the, some of the new models that have come out, you know, in the past, uh, couple of months now.

And so once we saw that, once we actually saw the code gen from it, we were like, "Oh my God." Like, okay, we can, we can build a product around this. And, and so that was really the impetus of us, of us building the thing.

But with that it was, you know, Stackblitz, the core Stackblitz product the past seven years has been an IDE for developers. So the entire user experience flow we've built up just didn't make sense. And so when we kind of went out to build Bolt, we just thought, you know, if we were inventing our product today, what would the interface look like given what is now possible with AI code gen?

And so there's definitely a lot of conversations we had internally, um, but you know, just kind of when we logically laid it out, we were like, yeah, I think it makes sense to just greenfield a new thing and let's see what happens.

If it works great, then, you know, we'll figure it out. If it doesn't work great, then we'll-- it'll get deleted at some point, right? Um, but, uh, but yeah. Anyway, so that, that, that's kinda how, how it actually came to be.

I'll mention your background a little bit. You, you were also founder of Thinkster before-

Yep

... you started Stackblitz. So both of you are second time founders. Both of you have sort of refounded your company recently. Well, yours was more of a rename.

Right.

I think a slightly different direction as well, and then we can talk about both. Maybe just chronologically, should we get caught up on, on where Qodo is, uh, first and then, you know, just like what, what people should know since the last pod?

Sure. Like the last pod was two months after we launched it, and we basically had the vision that we talked about the idea that, uh, software development is about specification, test and code, et cetera.

Yep.

And we are more on the testing part, as in essence we think that, uh, if you solve testing, you solve software development.

The, the beautiful chart that we'll put up on screen.

Yeah, exactly. That, that chart. And testing is, is a really big field. Like, uh, there are many dimension, unit testing, the level of the component, how, how big it is and how, how large it is. And, and then there's like different type of testing.

Is it regression or smoke or whatever. So back then we only had like, uh, one ID extension with unit tests as and, and focus. One and a half year later, first ID extension, uh, supports more, more type of testing as context to where we index, uh, local con-- local repos, but also ten thousands of repos, uh, for Fortune 500 companies.

We have another, uh, agent, another tool that is called-- the peer agent is the open source, and the commercial, uh, one is Qodo Merge. And then we have another open source called the Cover Agent, which is not yet commercial product, coming very soon.

It's very impressive. It could be that already people are approving, uh, automated, uh, pull requests that they don't even aware in really big open sources. So once we have enough of these, we will also, uh, launch, uh, another, another agent.

In one and a half year, uh, what we did is grew in our offering and mostly on the side of does this code actually works, testing, code review, et cetera. And we believe that's the, you know, the, the critical milestone that needs to be achieved to actually have the AI engineer, uh, for enterprise software, so.

And then, uh, like for the first year was everything bottom up, getting to one million installation. Uh, twenty twenty-four-- that was twenty twenty-three. Twenty twenty-four was starting to monetize to feel like, uh, how it is to, to make the first buck.

So we did the Teams offering. It went well with a thousand of teams, et cetera. And then, uh, we started like just a, a few months ago to do enterprise, uh, with everything you need, which is a lot of things that, uh, discussed in the last, uh, post that was just, just released, uh, by Codium.

Yep.

Uh, that's how we call-

Yeah

... our inner Codium. Uh, just, uh, opening the brackets, uh, our company name was, uh, CodiumAI.

Yeah.

And we renamed to Qodo, and we call our model Codium. So back to, to my point, so we started enterprise motion and already have like a-- multiple like Fortune 100 companies. And then with that, we raised a series A of, uh, forty, uh, million dollars.

And, uh, what's exciting about it is that enables us to develop more agents. That, that's our focus. Like, I think it's very different. We're not coming very soon with an ID or something like that.

You don't wanna fork VS Code?

Yeah, maybe we'll fork Jet, uh, JetBrains or something just to be different.

I noticed that, you know, I think the, the promise of general purpose agents has kinda died. Like everyone is doing kind of what you're doing. There's Qodo Gen, Qodo Merge, and then there's a third one. What's the name of it?

Yeah. The R- the Qodo Cover-

Cover

... uh, which is like a commercial version of, uh, Cover Agent. It's coming soon.

Yeah. I see something similar with Factory AI also doing like, uh, droids. They all have special purpose doing things, but people don't really want general purpose agents, right? The last time you were here, we talked about AutoGPT, the, the biggest thing of 2023.

Yeah.

This year, not really relevant anymore. And I think it's r- mostly just 'cause like I think when you give me a general purpose agent, I don't know what to do with it.

Yeah.

Like maybe not-

I, I totally agree with that. We, we're seeing it for a while, and I think it will stay like that despite the computer use, et cetera, that supposedly can just replace us and it could-- you can just like prompt it to be, "Hey, now be a QA," you know, or, "Be a QA person or, or a developer."

I still think that there's a few reasons why you see like, uh, dedicated agent. Again, I'm a bit more focused, uh, like my head is more on, you know, complex software or for big teams and enterprise, et cetera.

And even, you know, think about, uh, permissions and, uh, what are the data sources and just, uh, the same way you, you manage permissions for, for users. Like developers, you probably want to have dedicated guardrails and dedicated approvals for, for agents.

I intentionally like touched a point, um, not many people think about. And of course, then what you can think of, like maybe there's, uh, uh, different tools, tool use, et cetera. But just the first point by itself, uh, is a good reason why you wanna have different agents.

Just to compare that with Bolt.new, you're almost focused on like the application is very complex, and now you need better tools to kind of manage it and build on top of it. On Bolt, it's almost like, like I was using it the other day.

There's basically like, hey, like I'm just trying to get started, you know? I'm not very opinionated on like how you're gonna implement this. Like this is what I wanna do, and you build a beautiful... But what people ask as the next step, you know, going back to like the general versus like specific, have you had people say, "Hey, you know, this is great to start, but then I want a specific Bolt.new.whatever else to do a more vertical integration and kinda like development," or what's the-- what do people say?

Yeah, I think, um, I, I think you kinda hit the, hit it head on, which is, you know, kind of the way that we've, we've kinda talked about it internally is it's like people are using Bolt to go from like 0.0 to 1.0.

Like that's like kinda the biggest unlock that Bolt has versus most other things out there. I mean, I think that's kind of, you know, what's, what's very unique about Bolt. I think the, uh, you know, the working on like existing enterprise applications is, I mean, it's crazy important because, you know, there's, uh, you look, when you look at the Fortune 500, I mean, these code bases, some of these have been around for, you know-

Mm-hmm

... 20, 30 plus years. And so it's important to be going from, you know, 101.3 to 101.4, et cetera. I think for us, so what's been actually pretty interesting is we see there's kinda two different users for us that are coming in, and it's very distinct.

It's like people that are developers already, and then there's people that have never really written software. And, um-

Mm-hmm

... or if they have, it's been very, very minimal. And, and so in the first camp, what these developers are doing, like to go from 0 to 1, they're coming to Bolt, and then they're ejecting the thing to GitHub or just downloading and-

Mm-hmm

... you know, opening Cursor, like whatever, to, to, you know, keep iterating on the thing. And sometimes they'll bring it back to Bolt to like add in a huge piece of functionality or something, right? But for the people that don't know how to code, they're actually just, they, they live in this thing.

And that was one of the weird things when we launched is, you know, within, uh, a day of us being online, one of the most popular YouTube videos, and there's been a ton since, which is, you know, there's like, "Oh, Bolt is the Cursor killer."

And, and I originally saw the headlines- ... and I was like, "Thanks for the views?" I mean, I don't know. This doesn't make sense to me. That's not, that's not what we kinda thought.

This is how YouTubers talk to each other.

Well, but-

Everything kills everything else.

Totally. But what, what, what blew, what blew my mind was that there was any comparison 'cause it's like Cursor is a, is a local-

Right

... IDE product. But when, when we actually kinda dug into it and we t- and, and we have, uh, you know, people that are using our product saying this, "I'm not using Cursor now." We're like, "What?" And it turns out there are hundreds of thousands of people that we have seen that were using Cursor and were trying to build apps with that, where they're not traditional software devs, but were heavily leaning on the AI.

And as you can imagine, it is very complicated, right, to do that with Cursor. So when Bolt came out, they're like, "Wow, this thing's amazing," because it kinda inverts the complexity where it's like, you know, it, uh, it's not an IDE.

It's a, it's a chat-based sort of interface that we have. So yeah, so th- that's kinda the split, which is rather interesting. We've had like the first startups now launch off of Bolt entirely where this, you know, um, tomorrow I'm doing a live stream with this guy named Paul who he's built an entire CRM using this thing, and, you know, with backend, et cetera.

And, um, you know, people have made their first money on the internet, period, you know, launching this with Stripe or whatever have you. So, um, anyway, so that's, that's kinda the two main, the two main categories of folks that we see using Bolt, though.

I agree that I don't understand the comparison. Doesn't make sense to me. I think like we have like two type of families of tools. One is like we reimagine the software development. I think Bolt is, is there, and I think like, uh, Cursor is more like a evolution of what we already have.

It's like taking the IDE, and it's, it's amazing, and it's, uh, okay, let's, let's adapt, adapt the IDE to a era where LLMs can do a lot for us. And Bolt is more like, okay, let's rethink everything totally.

And I think we see a, a few tools there, like maybe, uh, Vercel, VO, and maybe Replit in, in that area. And then in the area of, uh, let's expedite, uh, let's change, uh, let's, let's progress with what we already have.

You can see Cursor and, and, uh, Qodo. We're, we're different, uh, between ourself, uh, Cursor and Qodo, but definitely I think like comparison doesn't make sense.

And just to set the context, this is not a Twitter demo. You made $4 million of revenue in four weeks. So this is, uh, this is actually working. You know, it's not a-

Yeah.

What, what do you think that is? Like, there's been so many people demoing coding agents on Twitter, and then it doesn't really work.

Yeah.

And then you guys were just like, "Here you go. It's live. Go use it." "Pay us for it," you know?

Yeah.

Is there anything in the development that was like interesting and maybe how that compares to building your own agents?

We had no idea, honestly. Like we, we've, we've been pretty blown away and, and things have just kinda continued to grow faster since then, where like, uh, today is week six. So I, I kind of coming back to the point you just made, right, where it's you, you kind of outlined i-it's like there's kinda this new market of like- Kind of rethinking software development, and then there's heavily augmenting existing developers.

I think that, you know, both of which are, uh, you know, AI, uh, code gen being extremely good, it's allowed existing developers it's, it's allowing existing developers to hammer out software far faster than they could have ever before, right?

It's like the ultimate power tool for an existing developer. But this code gen stuff is now so good, and, and we saw this over the past, you know, from the beginning of the year when we tried to first build Bolt.

It's actually lowered the barrier to people that, that aren't traditionally software engineers. But the kinda the key thing is if you kinda think about it from, imagine you've never written software before, right? My co-founder and I, he and I grew up, uh, down the street from each other in Chicago.

We learned how to code when we were 13 together, and we, we've been building stuff ever since. And this is back in, like, the mid-2000s or whatever. You know, there, there was nothing for free to learn from online on the internet on how to code.

So, like, we, for our 13th birthdays, we asked our parents for, you know, O'Reilly books 'cause you couldn't get this at the library, right? And so, um, instead of, like, an Xbox, we got, you know, programming books. But the hardest part for everyone learning to code is getting an environment set up locally.

You know, and so when we built Stackblitz, like, kinda the, the key thesis, like, seven years ago, the, the insight we had was that the, hey, it seems like the browser has a lot of new APIs, like WebAssembly and service workers, et cetera, where you could actually write an operating system that ran inside the browser that could boot in milliseconds.

And, and you would, you know, basically there's this missing capability of the web. Like, the web should be able to build apps for the web, right? You should be able to build the web on the web. Every other platform has that.

Visual Studio for Windows, Xcode for Mac. The, the web has no built-in primitive for this, and so just, like, our built-in kind of, like, nerd instinct on this was like, "That seems like a huge hole, and it's, you know, will be very valuable," or, like, you know, a very valuable problem to solve, so people have to set up dev environments.

You know, this is what we spent the past seven years doing, and the reality is existing developers have, running locally, they already know how to set up dev environments, so the problem isn't as acute for them. When we put Bolt online, we took that technology called WebContainer and married it with these, you know, state-of-the-art frontier models.

And the people that have the most pain with getting stuff set up locally is people that have, don't code. I think that's been, you know, really the, the big explosive reason is no one else has been trying to make dev environments work inside of a browser tab be- you know, for the past, uh, you know, since ever other than basically our company, largely 'cause there wasn't an immediate demand or need.

So I think we kind of find ourselves at the right place at the right time. And again, for this market of people that don't know how to write software, you would kind of expect that you should be able to do this without downloading something to your computer in the same way that, hey, I don't have to download Photoshop now to make designs 'cause there's Figma.

I don't have to download Word because there's, you know, Google Docs. They're kinda looking at this as, as that sort of thing, right? Which was kind of the, you know, our impetus and kinda vision from the get-go. But, uh, you know, the code gen, the AI code gen stuff that's come out has just been, you know, an order of magnitude multiplier on how magic that is, right?

So anyways, that's kinda my best distillation of, like, what, what is going on here, you know?

Yeah. And you can deploy too, right?

Yes.

In Stackblitz. Yeah.

Yeah, and so that's what's really cool is it, it's, you know, we have deployment built in, um, with Netlify, and, um, this is actually, I think, uh, Sean, you actually built this at Netlify-

Yeah

... when you were there. One of the most brilliant integrations actually because, you know, effectively the API that Sean built, maybe you can speak to it, but, like, as a provider, we, we can just effectively give files to Netlify without the user even logging in, and they have a live website.

If they want to keep, uh, hold onto it, they can click a link and claim it to their Netlify account. But it basically is just this really magic experience 'cause when you come to Bolt, you say, "I want a website."

Like, my mom, 70, 71 years old, made her first website, you know, on the internet two weeks ago, right? It was about her nursing days. And, and so, and it made up some-

Oh, that's fantastic though. It wouldn't have been made.

100%. 'Cause even in, you know, when w- we've had a lot of people building personal, like, deeply personal stuff. Like, in the first week we launched this, uh, this sales guy from the East Coast, uh, you know, replied to a tweet of mine, and he said, "Thank you so much for building this, you know, to, to your team."

His daughter has a medical condition, and so for her to travel, she has to, like, line up donors or something, you know, so ahead of time. And so he actually used Bolt to make a website to do that, to actually go and send it to folks in the, the region she was gonna travel to ahead of time.

I was really touched by it, but I also thought, like, why, you know, why didn't he use, like, Wix or Squarespace, right? I mean, th- this is, this is a solved problem, quote, unquote, right? And then when I thought...

I actually used Squarespace for my, for my, uh, the wedding website for my wife and I, like, back in 2021, so I'm, I'm familiar. You know, it was, it was faster. I know how to code, but I was like, "This is faster," right?

And I thought back, and I was like, there's a whole interface you have to learn how to use, and it's actually not that simple. There's, like, a million things you can configure in that thing. When you come to Bolt, there's an, there's a text box.

You just say, "I need a, I need a wedding website. Here's the date. Here's where it is." And, and here's a photo of me and my wife. Put it somewhere relevant. It's actually the simplest way, and that's what my, when my mom came, she said, "Uh, I'm Pat Simons.

I was a nurse in, in the '70s, you know, and, like, here's the things I did." And a website came out. So coming back to why is this such a, a... I think why are we seeing this sort of growth, it's this is the simplest interface I think maybe ever created to actually build and de- deploy a website.

'Cause and, and then f- that website my mom made, she's like, "Okay, this looks great." And there's, there's one button. You just click it, deploy, and it's live. And you can buy a domain name, attach it to it, and, you know, it's as simple as it gets.

It could get even simpler with some of the stuff we're working on. But anyways, so that's, uh... I- it's, it's, uh, it's been really interesting to see some of the usage like that.

I can offer my perspective. So I, you know, I, I probably should have disclosed a little bit that, uh, I'm a, a Stackblitz investor.

Cancel the episode.

From the... I know, I know.

Can't play it now. Uh, pause.

Uh, Eric actually reached out to show me Bolt, uh, before the launch, and, uh, we, you know, we talked a little, a lot about, like, the framing of, of, uh, what we were gonna talk, um, how we marketed the thing, but also, like, what Bolt was going to need, like, a, a whole sort of infrastructure.

Netlify, I was the maintainer of the CLI, but I won't take claim for the anonymous upload. Uh, that's actually the origin story of Netlify. We can have Matt Biilmann talk about it, but, um, he, he... Like, that was how Netlify started.

You could drag and drop your, a zip file or folder from your desktop onto a website. It would have a live URL with no sign-in. And so that was the origin story of Netlify, and it just persisted till today, and it's just like, uh...

It's really nice, interesting that, uh, both Bolt and CodeNation, Devin, and a bunch of other sort of agent-type startups, they all use Netlify to deploy because of this one feature. They don't really care about the other features of Netlify.

Yeah.

But, but just because it's easy for computers to use and talk to. Like, if you build an interface for computers specifically that is easy for them to navigate, then they will be used in agents, and I think that's a, a learning that a lot of developer tools companies are having.

That's my Bolt launch story and Netlify CLI stuff. Uh, and yeah, I just wanted to come back to, like, the web containers things, right? Like, uh, I think you put a lot of weight on the technical moats. Uh, I think you also are just, like, very good at product, so you've, you've, like, built a better agent than a lot of people, the rest of us, including myself, who have tried to build these same things, and we didn't get as far as you did.

Don't shortchange yourself on products. But I think specifically on, on infra, on, like, the sandboxing, like, this is a thing that people really want. Alessio has backed E2B, which we'll have on at some point talking about, like, the sort of the server full side.

But yours is, you know, inside of the browser, serverless. Yeah, you... It, it doesn't cost you anything to serve one person versus a million people. It doesn't, doesn't cost you anything. I think that's interesting. I think in theory we should be able to, like, run tests, 'cause you can run the full backend.

Like, you can run Git, you can run Node, you can run, uh, maybe Python, uh- ... someday. We, we talked about this. But ideally you should be able to have a full agentic loop, running code, seeing the errors, correcting code, and just kind of self-healing, right?

Like, I, I mean, is, isn't that the dream?

Totally.

Yeah.

Totally. At least in Bolt we've got, we've got a good amount of that, uh, today. I mean, there, there's a lot more for us to do, but one of the nice things... 'Cause, like, in, in WebContainer, you know, there's a lot of pr- kind of stuff.

You go Google, like, uh, you know, "Turn Docker container into WASM," like, there's, there... You'll find a lot of, uh, stuff out there that will do that. The problem is it's very big, it's slow, um, and then that ruins the experience.

And so what we ended up doing is just writing an operating system from scratch that was just purpose-built to, you know, run in a browser tab. And the reason being is, you know, these Docker to WASM things will give you a image that's like, oh, 60 to 100 mega- you know, maybe more, you know?

And our, our O- OS, you know, kinda clocks in I think, I think we're at, like, a, uh, maybe, maybe a megabyte or less or something like that. I mean, it's, it, it's, you know, really, really, you know, stripped down.

This is all... Basically f- the task involved, as I understand it, it's mapping every sing- single Linux sys call to some kind of Web, WebAssembly implementation.

But more or less. And, and, and then there's a lot of things actually, like when you're looking at a dev environment, there's a lot of things that you don't need that a traditional OS is gonna have, right? Like, you know, uh, audio drivers or, you know, like, there's just, like, there's just tons of things, right?

Oh, yeah, Decos.

Yeah, you just, you can just kinda, you can, you can kinda toss them. Or alternatively what you can do is you can actually be c- the, the, the nice thing... And this is, this kinda comes back to the origins of browsers, which is, you know, there, there...

At the beginning of the web in, you know, the late '90s, there was, you know, two very different kind of visions for the web, where Alan Kay vehemently disagreed with the idea that it should be document based, which is, you know, what Tim Berners-Lee, you know, that.

The, the... And that's kind of what ended up winning, winning, was this document based kind of browsing documents on the web thing. Alan Kay, he's got this, like, very famous quote where, you know, he said, you know, you, you want web browsers to be mini operating systems.

They should download little mini binaries and execute with, like, a little mini virtualized operating system in there. And what's kind of interesting about the history, not to geek out on this aspect, but what's kind of interesting about the history is and of...

both of those folks ended up being right. Documents were actually the, the pragmatic way that the web was... you know, became the most ubiquitous platform in the world, to the degree now that this is why WebAssembly has been invented, is that we're doing, we need to do more low-level things in a browser.

Same thing with WebGPU, et cetera. And so all these APIs, you know, to build an operating system came to the browser, and that was actually the realization we had in 2017, was holy heck, like, you can actually... You know, service workers, which were designed for allowing your app to work offline, that was the kind of the key one where it was like, wait a second, you can actually now run web servers within a browser.

Like, you can run a server that you open up. That's wild. And no-

Like full Node.js.

Full Node.js. Like, like, that capability. Like, I can have a URL that's programmatically controlled by a web application itself. Boom. Like, the web can build the web. The primitive is there. Everyone at the time, like, we talked to people that, like, worked on, you know, uh, Chrome and V8, and they were like, "Uh," you know?

"I don't know." So, but it's one of those things you just kinda have to go do it to find out. So we spent a couple of years, you know, working on it and, and, and, and got it to work in, back in 2021 is when we, uh, kind of put the first, like, beta of WebContainer online.

But, um-

In partnership with Google, right? Like, Google actually had to help you get over the finish line with some stuff.

Yeah, they're 100%, because what, you know, over the years of when we were doing the R&D on the thing- Kind of the biggest challenge, the two ways that you can kind of test how powerful and capable a platform are, the two types of applications are, one, video games, right?

Because they're just very compute-intensive, a lot of calculations that have to happen, right? The second one are IDEs, because you're talking about actually virtualizing the actual runtime environment you are in to actually build apps on top of it, which requires sophisticated capabilities, a lot-- access to, you know, a good amount of compute power, right, to, to effectively, you know, building app in app sort of thing.

So those, those are the stress tests. So if, if your platform is missing stuff, those are the things where you find out. Those are-- Those are the people building games and IDEs, they're the ones filing bugs on operating system level stuff, and for us, browser level stuff.

And so, yeah, what ended up happening is we were just hammering, you know, the Chromium bug tracker, and they're like, "Who are these guys?"

Yeah.

And, um, and they, they were amazing 'cause, I mean, just making Chrome DevTools be able to, to debug... I mean, it's, it, you know, it's not-- it wasn't originally built, right, for debugging an operating system, right? They've been phenomenal working with us and just kinda really pushing the limits.

But that, that-- it's a rising tide that's kind of lifted all boats because now there's a lot of different types of applications that you can debug with Chrome DevTools that are run in a browser that runs more reliably because just the stress testing that, that we and, you know, uh, games that are coming to the lab are kind of pushing as well.

But...

That's awesome. About, about the testing, I think, like, most, let's say, coding assistant from different kinds will need this loop of testing and even I would add code review to some, to some extent that you mentioned.

How is testing different from code review?

Code review could be, for example-

Just peer review?

Peer review, like, a code review that is done at the point of when you want to merge, uh, co- branches. But, uh, I would say that code review, for example, checks, uh, best practices, maintainability, and so-so on. Testing-

It's not just like CI, but-

Right

... it's more than CI.

And test- and testing is like, uh, more like checking functionality, et cetera.

Yeah, yeah.

So it's, it's different. We, we call, by the way, all of these together code integrity, but that's a different story. Just to go back to the, to the testing and, uh-

We're still here

... specifically-

Yeah. It's, it's, it's since the first-

Yeah. So-

... slide

... yeah, we're consistent. So if we go back to the testing, uh, I think, like, it's not surprising that for us, testing is important, and for Bolt, it's testing important, but I wanna, uh, you know, shed some light on a different perspective.

If like, let's think about autonomous driving. Those startup that are doing autonomous driving for highway and autonomous driving for the city. And I think, like, we saw the autonomy of the highway much faster reaching to level, I don't know, four or so, much faster than those, those in the city.

Now, in both cases, you need testing and, uh, quote, unquote, "testing," you know, verifying validation that you're doing the right thing on, on the road and you're reading and et cetera. But it's probably like, uh, so different in the city that it could be like actually different technology.

And I claim that we're seeing something similar here. So when you're building the next Wix, and I-- if I was Dan, I was, like, looking at you and being a bit scared, that's what you're disrupting, which is what you just said.

Then basically, I would say that, for example, the UX/UI is freaking important and, uh, because you're, you're more aiming for the end user. In this case, maybe it's an end user that doesn't know how to develop. For developers, it's also important, but let alone those that do not know to develop, they need a slick UI/UX.

And, and I think, like, that's one reason, for example, I, I think Cursor have, like, really good technology. I don't know the underlying, uh, what's under the hood, but at least what they're saying. But I think also their UX/UI is great.

It's a lot because they, they did, uh, their own, own ID. While if you're aiming for the, the city AI, suddenly, like, there's a lot of testing and code review technology that it's not necessarily, like, uh, that important.

Uh, for example, uh, let's talk about integration test. Like, probably like a lot of what you're building in Bolt at the moment is, uh, isolated, like, uh, uh, applications. Maybe the, the vision or, or the end game is maybe like having one solution for, for everything.

Like, it could be that eventually the highway AI, uh, s- uh, uh, companies, uh, will go into the city and, and the other way around. But at the beginning, there is, there is a difference, and integration tests are, are a good example.

I guess they're a bit less important, and when you think about enterprise software, they're really important. So to recap, like, I think, like, the idea of looping and, and verifying your test and, uh, verifying your code in different ways, testing or, or code review, et cetera, seems to me important in, like, in the highway AI and the city AI, but in different ways and different, uh, cr- uh, like, uh, critical for the city, even more and more va-variety.

Actually, I, I was looking f- to ask you, like, uh, uh, like, uh, what kind of loops you guys are doing. For example, when I'm using Bolt and I'm enjoying it a lot, then, um, I do see, like, uh, sometimes you're trying to catch the errors and, and fix them.

And also, I noticed that you're break-breaking down tasks into smaller ones and then et cetera, which is already a common, like, uh, notion for, for a year, a year from, a year ago, but it seems like you're doing it really well.

So if you're willing to share anything about that.

Yeah, yeah, totally. Yeah, yeah, yeah. I realize I never actually hit the punchline of, of, uh, what I was saying before. I mentioned the point about us kind of writing an operating system from scratch, because what ended up being important about that is that, to your point, it's actually a very...

Like, compared to like a, you know, if you're, like, running Cursor on a, a, anyone's machine, you kinda don't know what you're dealing with with the OS you're running on. There could be a, an error happens. It could be like a million different things, right?

There could be some config, there could be-- it, it could be God knows what, right? The thing with WebContainer is because we wrote the entire thing from scratch, it's actually a unified image basically, and we can instrument it at any level that we think is gonna be useful, which is exactly what we did when we w- started building Bolt, is we instrumented stuff at like the process level, at the runtime level, you know, et cetera, et cetera, et cetera.

Stuff that would just be n- not impossible to do on local, but to do that in a way that works across any operating system, whatever, is, I mean, would just be insanely, you know, insanely difficult to do right and reliably.

And that's what you saw, uh, when you've used Bolt, is that when an error actually will occur, whether it's in the build process or the actual web application itself is failing or anything kind of in between, you can actually capture those errors and, and, and today it's a very primitive way of, of, uh, how we've implemented it largely 'cause the product just didn't exist 90 days ago, so we're like, we got some work ahead of us and, and we gotta hire some more a little bit.

But, but basically we, we, we presented, we say, "Hey, this is-- Here's kind of the things that went wrong." Here- there's a fix it button and then a ignore button, and then you can just hit fix it, and then we take all that telemetry through our agent, uh, you know, run it through our agent and say kinda, "Here's the state of the application.

Here's kind of the errors that we got from Node.js or the browser," or whatever, and like da, da, da, da. And, and it can take a crack at actually solving it, and it's actually pretty darn good at, at being able to do that.

That's kinda been a, you know, closing the loop and having it be a, a, a reliable kind of base has seemed to be a pretty big upgrade over doing stuff locally just 'cause, um, I think that's a pretty key ingredient of it.

And yeah, I think breaking things down into smaller tasks, like that's, that's kind of a key part of our agent and I think like Claude did a really good job with artifacts and I think, uh, you know, us and kinda everyone else has, has kinda taken their approach of like actually breaking out certain tasks in a certain order into, you know, kind of a, a concrete way.

And, and so actually the, the core of B- uh, Bolt.new we actually made open source. So you can actually go like go and, and check out like the system prompts and et cetera, and you can run it locally and whatever have you.

So anyone that's like interested in this stuff, I'd highly recommend, uh, taking a look at. There's not a lot of like, uh, stuff that's like open source in this realm and so that was one of the fun things that we've, uh, we thought would be cool to do and, uh, people, uh, people seem to like it.

I mean, there's a lot of forks and people adding different models and stuff, so it's been cool to see.

Yeah. I'm happy to add, uh, I, I added real-time voice for my OpenAI Dev Day demo and, uh, it was really fun to hack with, so thank you for doing that.

Yep. Thank you, Itamar.

You didn't have to.

I'm gonna steal your code back, Itamar.

Yeah.

'Cause I want that mode.

It's funny 'cause I, I built on top of the fork of Bolt.new that already has the multi-LLM thing and so you just told me you're gonna merge that in so then you're gonna merge two layers of forks down- ...

into this thing, so it'll be fun.

Heck yeah.

Just to touch on like the environment, Itamar, you maybe go into the most complicated environments that even the people that work there don't know how to run. How much of an impact does that have on your performance? Like, you know, is most of the work you're doing actually figuring out environment and like the libraries?

Because I'm sure they're using outdated version of languages, they're using outdated libraries, they're using forks that have not been on the public internet before. How much of the work that you're doing is like there versus like at the LLM level?

One of the reason I, I, I was asking about, uh, uh, you know, what are the steps to, to, to break things down because it really matters, uh, like what's the tech stack, uh, how complicated the software is.

It's hard to figure it out when you're dealing with the real world, uh, any environment of enterprise. It's a city when I'm like, uh, while maybe sometimes like, uh, I think you, you do enable like, um, in Bolt like to install stuff, but it's quite of like a controlled environment and, uh, that's, that's a good thing to do because then you narrow down and, and it easier to, to make things, uh, work.

So definitely there are two dimensions. Uh, I think you-- Actually spaces. One is, is the fact just like installing our software without yet like, uh-

Doing anything

... uh, m- making it work.

Yeah.

Just installing it because we're, we work with enterprise and Fortune 500, et cetera, many of them want on-prem solution. So-

You have how many deployment options?

Basically we, we had a, we did a metric, uh, metrics, uh, to say 96 options- ... because y- you know, there are different dimensions. Like for example one dimension, uh, w- we connect to, to your, your code management system to your Git.

So are you having like a GitHub, GitLab, uh-

Subversion

... uh, is, is it like on cloud or deployed, uh, on, on-prem? Just an example. Which model we to use is APIs or ours, like we have our own models.

Is it TestGPT or something you c-

Yeah.

We named it Qodo.

We started with TestGPT.

Yeah.

It was a huge, uh, mistake, uh, name. It was cool at back then, uh, but I don't think it's a good idea to name a model after someone else model. Anyway, that's, that's my opinion. So we got, uh, actually-

I'm interested in these learnings, like things that you changed your mind on.

Eventually when you're building a company, you're building a brand and, uh, you, you wanna create your own, own brand. By, by the way when I thought about Bolt.new, I also thought about if it's not a problem because when I think about Bolt I do think about like a couple of companies that are already called this way.

Cursed companies.

Yeah.

You could call it Qodium just to...

Okay. Thank you. Tou- touche. Touche.

Yeah, you gotta imagine the, the board meeting before we launched Bolt, one of our investor-- You can imagine the, they're like, "Are you, you know, are you sure?" 'Cause from the investment side-

Yeah

... there was your kind of a famous, uh, very, you know, notorious Bolt and they're like, "Are you sure you wanna go with that name?"

Oh, yeah. Yeah, absolutely. At this point we have actually four models. There is a, a model for auto-complete, there is a model for the chat, there is a model dedicated for more for code review, and there is a, a model that is for code embedding.

Actually if, if... You might notice that there is not, there isn't a good code embedding model out there.

Mm.

Like c- can you name one?

Hm.

Like that would dedicate it for code like, uh, yeah, I, I can't share more than that.

There's, there's code indexing and then you can do sort of like the hide for code and then you can, you can embed the descriptions of the code.

Yeah. But, uh, you do see a lot of, uh, type of models that, that are dedicated for embedding and for different space and different, uh, like, uh fields, et cetera and I, I, I'm not aware like, uh- And I know that, uh, th- if you go to the Bedrock, try to find, like, there is a few co-embedding models, but none of them are specialized for, for code.

Is there a benchmark that you would t-tell us to pay attention to?

Yeah. So i-it's coming.

Ah.

Wait for that. Anyway, like, we have our, we have our models, and just to, to go back to the, uh, the opt- the ninety-six option of deployment-

Yeah, sorry, I sidetracked you.

So, yeah, so, uh, in closing the brackets for us. So, um, so one is, like, dimensional, like, what Git deployment you have. Docker, like, what models are-- do you agree to use? Docker, uh, could be, like, if it's air-gapped completely or you want a VPC, and then you have Azure, GCP, and, uh, and AWS, which, which is different.

Do you use Kubernetes or do not? Like, because we want to exploit that. Uh, there are companies that do not do that, et cetera. And I guess you know what I mean. So that's, that's one thing. And considering that we are dealing with-- wanna solve for enterprise, we needed to deal with that.

So you asked me about, like, how complicated it is to, to solve that complex code. I said it's just the deployment part, and then, like, uh, n-now to the software, we see a, a lot of different, like, challenges.

For example, some companies, they did actually a, a good job to, uh, build a lot of microservices. Uh, let's not get to the flo-- if it's good or not, but let's, let's for a second assume that it is good, a good thing.

A lot of microservice, each one of them has their own repo, and now you have tens of thousands of repos, and you, as a developer, want to develop something and, uh... And I remember me for-- coming to a corporate for a first time, I don't know where to look at, like, uh, uh, where to find things.

So just, uh, doing a good indexing for that is, like, a, a challenge. And, uh, moreover, the regular indexing, the one that you can find... Well, we wrote a few blogs on that. By the way, we also, like, have some open source, like, different than yours, but actually three and, and growing.

Then it's not-- it doesn't work. You need to let the tech leads and the companies influence your indexing. For example, mark with dif- different, uh, repos with, with different colors. This is a high-quality repo. This is a lower quality repo.

This is a, a repo that we wanna deprecate. This is a repo we want to, to grow, et cetera. And let that be part of your indexing. And only then, like, things actually work for enterprise, and they don't get to a fatigue of, "Oh, this is awesome.

Oh, but I'm starting-- Like, it's annoying me." I, I, I think Copilot is an amazing tool, but I'm quoting others, meaning GitHub Copilot, that they see not so good retention of GitHub Copilot in, in enterprise.

Ooh, spicy.

I saw snapshots of people, like, uh-- And we have customers that, that are Copilot users as well. And, and also I saw, like, uh, research, uh, some of them is public, by the way.

Yeah.

Between thirty-eight to, uh, fifty percent retention for users using Copilot in, in enterprise.

Oof.

So it's, it's not, it's not so good, right? By the way, I don't think it's that bad, but it's, it's not so good. So I think that's a reason because, yeah, it helps you, like, auto-complete, but then... And especially if, if you're working on your repo alone, but if it's need that context of remote repos that your code base, that-that's hard.

So to make things work, there's a lot of work on that, like h- get it-- yeah, giving the controllability for the tech leads, for the developer platform or developer experience department in the organization to influence how things are, are working.

A short example, because already long, um, if you have, like, really old legacy code, probably some of it is, like, is not so good anymore. If you just fine-tune on these code base, then there is a bias to repeat those mistakes or old practices, et cetera.

So, like, you, you need, for example, as I mentioned, to enable like, uh, to influence that. For example, in Qodo, you can have a markdown of best practices by the tech leads, and, uh, Qodo will include that and relate to that and will not offer, like, suggestion that are not according to the best practices.

Just as an example, okay? So, like, that's just a short list of things that you need to do in order to deal with, like the, like you mentioned, the one hundred to one, one hundred point one to one hundred point, point two, uh, version of, of software.

I just wanna say, like, what you're doing is extremely impressive because it's very difficult. I mean, the business of Stackblitz kind of before Bolt came online was sell-- we sold a version of our IDE that w-went on-prem. So I understand what you're saying about the difficulty of selling on-- you know, getting stuff just working on-prem.

Yeah.

Holy heck. I mean, that, that is extremely hard. I guess the question I have for you is, I mean, we, we were just doing that with, you know, kind of Kubernetes-based stuff. The spread of Fortune five hundred companies that you're working with, how are they doing the inference for this?

Are they, are they just... A-are you kind of plugging into, you know, Azure's OpenAI stuff and AWS's Bedrock, you know, Claude stuff? Or, like, or are they just, like, running stuff on GPUs? Like, what is that like? How are these folks approaching that?

Because, man, what we saw on the enterprise side, I mean, I gotta imagine that th-that's a huge challenge.

Everything you said and more, like- It's an example. Like, someone could be and, uh, I don't think any of these is bad. Like, um, they, they made a decision. Like, for example, some people, they're, "I want only AWS and VPC on AWS no matter what."

And then they-- Some of them, like, there's a subset I will say, "I'm willing to take models only for-- from Bedrock and not, not ours." And we have a problem because there is no good code embedding, uh, model on Bedrock, and that's part of what we're doing now with AWS to solve that.

We solve it in a different way, but, but if you are willing to run on AWS VPC, but run your, uh, run models on, uh, on GPUs or Inferentia, like the new version of the more coming out, then our, our models can, can run on that.

But everything you said is, is right. Like, we see, like, uh, on-prem deployment where they, where they have their own, own GPUs. We see Azure where you're using OpenAI Azure. We see cases where you're running on GCP, and they want OpenAI.

Like this cross, like, uh, uh, case. Although there is Gemini or even Sonnet, I think is available on GCP.

Yeah.

Just an example. So all the all options, that, that's part of the, the challenge. I admit that we thought about it, but it was even more complicated, and-

Yeah

... it took us a f- a few months to actually, that metrics that I mentioned, to start, like, clicking each one of the, the blocks there.

I mean, a few months is impressive. I mean, honestly, just that's... Okay, and every one of these enterprises is, their networking is different. Uh, it just, everything's different.

Yeah.

Every single one is different. And, and they're-

As you, as you understand.

Yeah, yeah. So, so that just cannot be understated. Like, like, that it is, it, it, that's, I'm, uh, extremely impressive. Hats off, 'cause it's tough.

It could be, by the way, like for example, "Oh, we're only, uh, AWS. Ah, but our, our GitHub u- Enterprise is on-prem." "Oh, we forgot, so we need like a private link or whatever."

Yep.

Like, every time like that. It, it's, it's, it's not... And, and you do need to think about it if you want to work with a enterprise, and it's important. Like, I understand, like, uh, their... I respect their point of view.

And this primarily impacts your architecture, your tech choices. Like, you have to-- You can't choose some vendors because of-

Yeah, definitely. To be frank, it makes this hard for a startup because it means that we want, we want everyone to enjoy all the variety of models. By the way, it was hard for us with our, our technology, and, um, I wanna open a, a, a bracket, like a window.

I guess you're familiar with our AlphaCodium, which is an open source-

We gotta go over that. Oh, yeah.

Yeah. So I-

Okay

... will do that quickly. Um-

Yeah, a pin in that. Yeah

... ah, actually, we did have it in the last, uh-

Yeah

... episode. So, so, okay.

Okay, we'll come back to that later, but let's talk about-

Yeah. So, so just like shortly, and then we can double-click on, on AlphaCodium. But AlphaCodium is a open source tool, you can go and try it, and lets you compete on Codeforces as a, as a website and a competition, and actually reach a master level, level, like 95 percentile with a click of a button.

You don't need to do anything. And part of, uh, what we did there is, uh, taking a problem and build- breaking it to different, like, smaller blocks and, and then the models are much, doing much better job. Like, we all know it by, by now that d- taking small task and solving them.

By the way, even o1, which is supposed to be able to do system two thinking, like Greg, uh, from OpenAI, like, uh, hinted, is doing better on these kind of problems, but still, it's very useful to break it down for o1, despite o1 being able to think by itself, and that's what we presented like just a month ago.

OpenAI released that now they are doing 93 percentile with o1 IOI of the, um, International Olympiad, uh, um, mathe- information. Sorry, I forgot exactly. Uh, I told you I forgot. And, um, and we took their o1 preview with AlphaCodium and did better.

Like, it just shows, like... And, and there is a big difference between the preview and, uh, the IOI. It shows like that these models are not still system two thinkers, and there's a, a big difference. So-

Yeah. Maybe they're not complete system two.

Yeah.

Yeah, they need some guidance.

I, I call them system one, 1.5. We, we can- ... we can dive into that. I thought about it. Like, you know, I, I care about this philosophy stuff, and I think, like, we didn't see it even close to a system two thinking.

I can elaborate later. But closing the brackets, like, we take AlphaCodium and as, as, as our principle of thinking, we take, uh, tasks and break them down to smaller tasks, and then we want to exploit the best model to solve them.

So I want to enable anyone to enjoy o1 and Sonnet and Gemini 1.5, et cetera. But at the same time, I need to develop my own models as well because some of the Fortune 500 want to have all air-gapped or, or whatever.

So that, that's a challenge. Now you need to support so many models and, and to some extent, I would say that the flow engineering, the, the breaking down to, to different blocks, is a necessity for us. Why? Because when you take a big block, a big problem, you need a very different prompt for each one of the models to actually work.

But when you take a, like a big, a problem and break it into small tasks, we can talk how we do that, then the prompt matters less. What I want to say, like, uh, all this, like, as a startup trying to do different deployment, getting all the juice that you can get from models, et cetera, is a big problem, and one need to think about it, and one of our mitigation is that process of taking tasks and, and breaking them down.

That's why I'm really interested to know how you guys are doing it, and part of what we do is also open source, so you can see.

There's a lot in there. But yeah, flow over prompt, I, I do believe, uh, that that does make sense. I feel like there's a lot that both of you can sort of exchange notes on-

Yeah.

Um, breaking down problems-

Yeah

... and I just want you guys to just go for it. Whatever. Like, this, this is fun to watch.

Yeah, I mean, what's super interesting is the context you're working in is, is, uh... 'Cause, you know, for, for us too with Bolt, we've started thinking because, you know, our, our kind of existing business line was, you know, going, you know, behind the firewall, right?

We were like, "How do we do this?" Adding the inference aspect on, we're like, "Okay, how does..." 'Cause there, I mean, there's not a lot of prior art, right? I mean, like, this is all new, you know? This is all new.

Yeah.

So I definitely am gonna have a lot of questions for you.

I'm here.

Um-

We're very open, by the way. We have a paper on it and a blog, like whatever, like, uh-

Yeah, it's on the AlphaCodium, uh, GitHub, and, uh, we'll, we'll put all this in the show notes as well.

Yeah, and I... And even the new results of, uh, o1, we, we published it and-

I love that. And I, I also just, I think spiritually, I like your approach of being transparent and, you know, 'cause I think there's, there's a lot of hypeium around the AI stuff and, and a lot of it is it's just like you have these companies that are just trying to keep their stuff closed source and, and then just max hype it, but, but then it's, it's kinda nothing.

And, and I think it kinda gives the, a bad rep to the incredible stuff that's, that's actually happening here. And so I think it's, it's stuff like what you're doing where, I mean, true merit, and you're cracking open actual code for others to learn from and use.

You know, that strikes me as the right approach, and I, and, and it's great to hear that you're making such incredible-

I have something to share about the open source. Like most of our tools are, uh, we have like a open source version, and then like a premium pro version, but it's not an easy decision to do that. I actually wanted to ask you about your strategy, but I think, like, in your case there is, in my opinion, relatively a good strategy where a lot of parts of open source and...

But then you have the deployment and the environment, which is not, right? If I get it correctly.

Mm-hmm.

And then there's, there's a clear, like, almost hugging face, like model. Like, yeah, you can do that, but why should you try to deploy it yourself? Deploy it with us. But in our case, and I'm not sure you're not gonna hit also some, some competitors, and, and I guess you, you are.

I wanted to ask you, for example, on some of them In our case, like one day, uh, we looked on one of our competitors that is doing code review. We're a platform. We have like the code review, the testing, et cetera, spread all over the ID to get...

And in each agent we have a, a few startups or a big, big, uh, like, uh, incumbents that are doing only that. So we noticed one of our competitors having not only a very similar UI of our open source, but actually even our typo.

And you know, you sit there and, and you kind of like, uh, yeah, we, we were not that good. Like, we don't use enough like, uh, Grammarly whatever, and, uh- ... and we had like a cal- a couple of these and we saw it there and, and then it's a challenge.

Yeah.

Like, and I wanna ask you like, uh, you know, Bolt is doing so well, and then you open sourced it. So I think I know what, what, that my answer was. I gave it before, but still interesting to hear what, what you think.

Uh, Geo had said back, um, I think it, when he was, when he was, uh, I don't, I don't know if he's up to date at this exact moment, but yeah, I think when on comma.ai, you know, all that stuff's open source.

And someone had asked him like, "Why, why is this open source?" And he's like, "If you're not actually confident that you can go and crush it and, and build the best thing, then yeah, you should probably keep your stuff closed source."

Right?

Yeah.

Like that, he said something akin to that. I'm probably kind of butchering it, but I thought that was a really, it was kind of a really good point. And that's, and that's not to say that you should just open source everything, 'cause I, for obvious reasons, there's like kind of strategic things you have to kinda take in mind.

But I actually think a pretty liberal approach as, as liberal as you kinda can be, it can really make a lot of sense 'cause I mean, that is so validating that like one of your competitors is taking your stuff and they're like, "Eh," like, "Let's just, you know, kinda tweak the styles and you know."

I mean, clearly, right? I think it's kind of healthy 'cause it, it keeps... I, I'm sure back at HQ that day when you saw that, you were like, "Oh. All right, well let's... We, we gotta, we have to grind even harder, right?

To make sure we stay ahead." And so I think it's actually a very useful motivating thing for the teams, 'cause you might feel this period of comfort. I think a lot of companies will have this period of comfort where, you know, they're, they're, they're not feeling the competition, then one day they get disrupted.

So kind of putting stuff out there and letting people push it forces you to face reality soon, right? And, and actually feel that incrementally so you can kind of adjust course. And that's, for us, the, the open source version of Bolt has had a lot of features people have been begging us for, like persisting chat messages and checkpoints and stuff.

Within the first week, that stuff was landing in the open source versions and they're like, "Why can't you ship this? It's in the open, so people have forked it," and- ... and we're like, we're like, "We're trying to keep our servers and GPUs online."

Like we're-

Yeah, yeah.

You know, like we're just... And, but it's been great 'cause it's, um, you know, like the folks in the community did a great job, kept us on our toes and, uh, and, and there's, and we've, we've got to know most of these folks too at this point that, you know, have been building these things.

And so it actually was very instructive, like, okay, well if we're gonna go kind of land this, they, there's some UX patterns we can kind of look at and the, and the code is open source to this stuff.

You know, what's great about these, what's not. So anyway, so I think that's, um, you know, anyways, net net, I think it's, it's, it's awesome. I think from a competitive point of view for us, for us, I think in particular what's interesting is, is the, uh, you know, kind of the core technology of WebContainer going and I, I think that right now there, there's really nothing that's kind of on par with that.

And we also, like, we have a business of... 'Cause you know, WebContainer runs in your browser, but to make it work, it ha- you have to install stuff from NPM. You have to like make cores bypass requests, like connect to databases, which all require server side, you know, proxying or acceleration.

And so we actually sell WebContainer as a service. The original, one of the core reasons we open sourced kind of the core components of Bolt when we launched was that, you know, we think that there's gonna be a lot more of these AI, in your browser AI codegen experiences, kind of like what Anthropic did with Artifacts and Claude.

By the way, Artifacts uses WebContainers.

Yeah.

Not yet.

No, uh, yeah.

Should I strike that?

I think that they've got their own thing at the moment, but there's been a lot of interest in WebContainers from folks doing things in that sort of realm and, you know, sc- in, in the AI labs and startups and everything in between, you know?

And so I think there, there'll be, I imagine, a good, you know... Over the, over the coming months there'll be lots of things to being announced of folks kind of adopting it. But yeah, I think, I think effectively, uh ...

Well, okay, I'll say this. If, if you're a large model lab and you want, uh, you know, to build, uh, you know, sandbox environments inside of your chat app, uh, you should call Eric.

But, but wait, wait. Wait, wait, wait, wait, wait. Like, uh, I have a question about that.

What?

Like, I think like OpenAI, um, you know, they felt that people are not using their model as they would want to, uh, so they built ChatGPT. But I would say that ChatGPT now defines OpenAI.

Yeah.

I know they're, they're doing a lot business from their, their APIs, but still, is this like how you think? Like isn't Bolt.new is, is your, your business now? Like why, why don't you like put-

Ooh

... focus on that instead of the-

What's your advice as a founder?

Well, well, I'll, I'll... You're, you're right, and so, so what we're going into it, we candidly, we were like, okay, you know, Bolt.new, this thing is super cool and like we think people are stoked. Well, we think people will be stoked, but we're like maybe this'll add, you know, uh, uh, best case scenario after month one we'd be mind blown if we added a couple hundred K of ARR or something.

Okay.

You know? And, and we were like, but we think there's probably gonna be a, a, you know, an immediate huge biz- 'Cause we, there was some early poll on, uh, you know, folks wanting to put, put WebContainer into their product offerings, you know, kind of similar to what Bolt is doing or whatever.

We were actually prepared for the, the inverse outcome here. But uh, I mean, well, I guess we've seen poll on both, but I mean, what's happened with Bolt-

Yeah

... and you're right, it's actually the same strategy as like OpenAI or Anthropic where, you know, we have a, our, our, you know, ChatGPT to OpenAI's APIs is, you know, Bolt to WebContainer. And so that's, it, it, we've kind of taken that same approach and we're seeing, I guess, kind of some of the similar results except our, right now the, the revenue side is extremely lopsided to-

Yeah

... to Bolt, right? And so-

I, I think like, uh, if you ask me what's my advice, I think you have three options. One is to focus on Bolt, the other is to focus on the WebContainer, the third is to raise $1 billion and- ...

or whatever and do them both. I'm, I'm serious, like I think like otherwise you need to choose. And if you, if you raise enough money, like, and I think it's big bucks because you're going to be chased by competitors and, uh, it's, I think it'll be challenging like to do, to both.

Like, and maybe, maybe you can, I don't know. Like we, we do see these numbers right now raising above $100 million even without having a product.

Yeah.

Uh, you can see these.

It's excellent advice. And I, I think, I think what's been amazing but also kind of challenging as you, as you, you know, kind of we're, like, trying to, trying to forecast, you know, okay, well, what, where are these things going?

'Cause I mean, in the, in the initial weeks, I think us and all the, you know, investors in the company that we were kind of sharing this with, we were... It was like, "This is cool. Okay, we added 500K.

Uh, wow, that's crazy. Wow, we're at a million now." Okay. You kind of, you know... Most things, you have this kind of the, the TechCrunch launch of initiation, and then the, the thing of sorrow. And this is, this is just, you know, we haven't...

If, if there's gonna be a, a downtrend, it just- it's just not coming yet. Now that we're kind of looking ahead, you know, we've- we're six weeks in, so now we're getting enough confidence in our convictions to go, "Okay, this, this kind of seems to be, this seems to be the trend line and kind of how we should go."

I can tell you another reason why I think, like, uh, where, where is Jasper?

They actually just announced some new numbers recently. Like, they, you know, they're, they're still surviving. They have gone down a lot. I think the, the peak that I heard was 100 million ARR-

Right

... and then they, and then they died.

And now, you know, there's, like, tens of these.

Yeah.

So I think, like, um, the, the... And, and their success was phenomenal, like what I-

Yeah

... I see at Bolt. And, and I think, like, if, if you wanna keep that, probably it's... Who am I? Like, I'm just, like, giving my, uh, two cents.

No, I think you're right though. Yeah, yeah, yeah.

You need to focus because you are, like, you are going to see, like, uh, Wix, I think, like, you're disrupting their market, and you open sourced some of it. And, uh, they have containers, I believe.

Yep.

And you, you will need to fight. Like, I can tell you that when we open source, I, I share with you, like, a small competitor, but I can tell you I have a friend who's built a billion-dollar company, uh, and more.

Like, when we released AlphaCodium, he, like, sent me a private email asking, "What the fuck did you just do?" Like, "Why did you release that as-

Like, you should've kept it.

Yeah. Like, "You released that open source." Like, um, you know, I'm thinking build some stuff, and now I can do that much easy, much, much more easily.

Yeah.

I, I can tell you my answer, and I think, I thought that maybe you will answer as well, although, like, I think Bolt al- already very promising. For us, like, AlphaCodium-1 is like GPT-1. I agree with you, like, being open and open source, et cetera, really helps to improve the product, community, et cetera.

But at some point, OpenAI closed their GPT-3.5 or, or whatever.

Yeah.

And, uh, that was part of my answer. Like, AlphaCodium is the agent that is compatible w- with, uh, GPT-1, and there is a lot to do for these agents to actually get that, like, moment that we had with GPT-3.5, et cetera, as agents.

Yeah. I, I think you're dead right. And, and I think it, it just comes back to... It kind of comes back to what Gio Hot said, right? It's like, y- you know, if you wanna win, there's no other option than out-hustling everyone else, you know?

And, and so I think that's kind of... And, and then out-hustling in the sense really meaning building the best product, building the best experiences. And so... And I think, um, that's the only way, kind of almost any route.

And, um, it, you know, kind of open sourcing some of the stuff just kind of burns the ships in a sense, right? And then maybe that's the simplest way of saying it. Just you're, you're burning the ships, but it also, it g- it builds a lot of goodwill.

Yeah, I mean, there's a lot, tons of benefits to it, right?

Yeah. Salesforce are doing that, right? They're now gonna be Agentforce-

Agentforce

... or whatever, so you can also

We're, we're gonna try to get Mark, uh, on, on the podcast. And, uh, they're g- Decibel is good friends with, uh, Salesforce. Any parting thoughts? Any, you know, trends that you're, you're, like, super excited about?

If we're talking about trends, I go back to our original, uh, podcast where we talked about, like, the idea that the software world is devi- is, like, built from specs, tests, and, and code. And I think, like, you can see that, uh, like, one dimension are company startups that are rethinking the entire, like, uh, development, uh, environment.

Like, I think, like, uh, Bolt, et cetera. And another dimension is, like, where is their focus? Is it, like, on the spec? Is it on the test and on the code? And I think, like, it's interesting to see that from, from that view.

We'll see more startup and more amazing announcement, like, of new, you know, directions, new philosophy. So I think we'll see startup, like, focusing, "Let's build everything from the spec." To some extent, I would say that, like, Bolt is, from my understanding, you can say better, some- somewhere in the line between the spec and the code because you start...

Like, I saw your demos. You're trying to describe, like, things, like, not just in one row because you want, like, to look like, like you want in it. So it's on that edge between connecting bet- between spec and code.

And you see others, like, uh, like, I think, like, all the IDEs, most of them are, are, uh, the new IDEs with a fork are, are there. Uh, w- we are, like, uh, more focused from the test and to the code and to, to the spec, et cetera.

So these are the trends. I think, like, we will see that. And, uh, and I think, like, the, the, another dimension to, to consider is, like, is it more for the highway AI, for the developers, maybe not even, like, the technical person or, or, or is it for, for the enterprise?

And that also, like, gives you, like, different products if they are aiming for different ICP, different, uh, ideal client profile. They will approach this triangle of spec and test and, and code. And, and, and that's, like, how I see the world.

And what I'm noticing is that we're seeing more and more of those new startups, new interfaces that is, are not focused on code. For example, having, talking more about the spec, talking more about the testing. Eventually, I think that that's where the world is going to.

Like, uh, the code is gonna be there, and where there will be developers, et cetera. But as agent improves and capabilities of the LLMs and, and integrations to different parts of, of the, uh, environment, development environment, we're going to see more and more, like, focusing on the spec and, and the test.

Basically, like, these two might unite, the spec and the test, because you can say that tests are runnable specs right? To some extent. So, like, that's another way to look at it. Yeah. That's-

It's literally on the slide here, runnable tests-

Yeah

... right here.

Yeah. I'm consistent.

It's how consistent-

Yeah. No, no, no.

Uh, it's-

Look, look, I talked about System 1 and System 2, like, uh, more than a year ago.

Yeah, yeah.

And now, now-

It's coming

... with o1, people are talking about System 1.

We're halfway there.

Uh, System 1.

Yeah.

And, and, but, but I think we'll talk about it again because I think they're totally, totally wrong about o1 being a System 2. So, like, uh, it, it is now, like, in the hype or whatever talking about that, but I think, like, the agents are the one that will take-

Yeah

... us towards System 2. And the more they are aware of their environment and aware of that sometimes they don't know what they don't know-

Then we'll really get to system two. But that's like a deeper discussion than-

That's a deeper discussion. Uh, love the philosophy talk that we had last time as well. All right, so we're, we're, we're back onto Bolt and, uh, you know, uh, Itamar has to leave for, for another interview. But we were just talking about like the what happened post-launch, right?

Yeah.

And, um, you know, I held this like sort of emergency council of, of advisors for you- ... uh, because we had never seen this before and I was like, "Okay, I'm gonna call all the smartest people I know to, to, to join this thing."

Which was extremely helpful. And I, I am so appreciative. There's been a, a handful of me-

You made a, you made a couple- You made a, you made one hire out of, out of that.

Yeah, 'cause I, I, you know, it was like, I think I, I can't remember where we were at kind of ARR-wise, uh, when I'd, when I'd messaged you.

It was like, you messaged me at like 2:00 or 3:00.

Okay.

And then by the time we got everything together it was 4:00.

Four, yeah.

And then, yeah, now it's at-

Since Eric sat down, five minutes.

But, I mean, it, it sounds like it ac- accelerated 'cause you told me it was like 100K, 200K a day.

Yeah.

And now it's accelerated?

Yeah, the, I, this, this past... I mean, I, I think every week has been kind of a blowout week as far as- ... I, I-

Is it TikTok? Is it-

We are digging into the degree that we can of just like where all this stuff's coming from. I mean, there's a, a ton of word of mouth, right? So you can't, what you can't just like look by refer, right?

So there's a ton of direct. But yeah, I mean, the, the, there's a lot of TikTok, there's a ton of YouTube. It's kind of, I think, been a sensation in these sort of like, uh, you know, entrepreneurial build-your-own SaaS, indie hacker, you know, even developer circles.

And I think, too, like our team's been doing a really good job. Our folks just kind of like flipped a switch and like we, you know, people were just working through the weekends or whatever to get stuff fixed and, and so the product...

And you'll see people, people say this on there, online, like today there was a tweet, someone was like, "Yeah, I tried this like the first week and like couldn't get whatever to work. Came back today, you know, m- uh, six weeks later, and this is ridiculous.

Like, this is so good." Right? And so I think there's, there's been an incredible amount of improvement to the product, to the agent, also to like the underlying models, too. Like Sonnet, uh, you know, they just happened to do an update on, you know, uh-

Yeah. Yeah

... with their, with their release like a couple weeks ago. And, and so, you know, when we put our new agent online and the new Sonnet, we saw con- a huge bump in conversion just based on that. And so, anyway, so yeah, like, you know, we've, we've gone, I think when we were chatting, that must've been three weeks ago.

You know, we had maybe an average of 100K in ARR, about, per day. And, you know, this week I will see, I've said this every week, but we'll see if it holds. You know, it's like kind of we've, the past couple days have been like, you know, half a million of ARR per day, just, I mean, which, you know, insane.

I mean, uh, uh, you know, I think today we had our peak, we've had peak traffic, you know, just kind of set the previous... You know, and that's kind of been, you know, every day this week. But anyways, yeah, I think things, the things just continue to accelerate, which is kind of blowing my mind, you know, 'cause it's just the, the sheer numbers of the stuff are just like mind-boggling.

You know?

I, I think you, you almost suffered from like the Twitter demo issues that other people had. Like, the first time I saw Bolt, I saw the demo and I was like, "Oh, that's cool." Like, I didn't go to try it because I was like, "I've seen so many of these-"

Yeah, yeah

... "that it's like I don't know if it's actually gonna work."

Yeah.

And then two days ago I signed up to use it. I was building a Luma replacement. I, I'm done with Luma. And I was like, "Man, this thing really works." I was like... And I, I already knew you, of course.

And I was like, "Man, this thing really works. What the fuck?" I was like... It, it's actually, I don't know if it's like the model, if it's like how you prompt it, but it's so good at coming up with like the simplest thing to implement.

So the Luma example, right? So first I was like, "Create a RSVP page for an event," and it created a wedding RSVP. I don't know if that's your fault. I don't know if you built that into the model.

Um, and then I was like, "Well, now it needs to have a way to create more events," and it added that. And then I was like, "Now it needs a way to like have a admin page to modify event."

And maybe what I would've done as a developer is like, "Well, I'll create a different like admin view, you know, with all the events, and then I'll have like the front end thing." And instead what it did is like it created like a admin view toggle on top and then like just a pencil button on every page to edit them in line, you know?

And that was it. And I was like, "Yeah, that works just as well."

Yeah.

And like for the model, that's probably like the simplest way to do it because it like limits the amount of files that are there.

Yeah.

Can you talk just more about how much of this is like the model coming up with it, how much you're prompting it to kind of like be very like compressed and concise?

A ton of it is the model, but I think y- what's interesting, though, is, you know, kind of the baseline model if I, just to like, if to kind of like try and put it into like a, you know, a way if you had to quanti- quantify, you know, the effect is obvi- the model is like this sort of like 10X multiplier.

Your, how good the bottom line model is, huge, huge swing. And then kind of what you can do on top of that, you can squeeze out three, 4X kinda more. And so that's kind of where the realm of, you know, prompt engineering and a, you know, multi-agent approaches, et cetera, kind of kick in.

And so I, I think, I think with us, you know, our, our folks, like with the guy on our side that, um, you know, led up the web container engineering, like that kind of our core technology for the past, you know, seven years here, you know, his name's Do- Dominic Elm, based out of Germany, and, uh, he was one of the founding engineers of the company.

You know, previous to Stackblitz he actually was doing machine learning, and he basically had built a Stackblitz like online IDE for machine learning. So think like a kinda like Google Colab-

Mm-hmm

... sort of thing-

Yeah, yeah

... or like Hugging Face has their kind of version of this. Back in 2016 there wasn't as much of a market for this stuff.

Mm-hmm.

But anyways, so he, he had, he had been doing a lot of, you know, training, you know, ML models and that sort of thing. So anyways, you know, as we began, uh, you know, kind of digging into AI stuff, you know, over the past year, he's been kind of leading that off.

And so a, a lot of it I, you know, I really attribute to, to Dom's, uh, specific angle 'cause he has deep understanding of our technology and how it works 'cause he's, you know, led the engineering on WebContainer, but has a, you know, deep understanding of how these models work.

Going and actually kind of writing out these, uh, you know, whether it's like the, the, the prompt engineering aspect of it or multi-agent or whatever have you, you know, that's sort of like that much context. And then, and the other folks on the team are, are, you know, in the same, same sort of spot.

They've, um, been working on this stuff. You know, I think we've been able to squeeze out a lot more than I've seen almost anything else out there, at least in the term of building web apps at least. But anyways, I, I think it's, I think it's kind of just because we, uh, uh, we have, you know, more context on, on a fewer number of heads at the company, so we can kind of connect the dots a bit faster, you know?

Yeah. That's part of the issue with the whole raise a billion dollars thing, like you actually run very lean.

And that's, that's actually been to your advantage.

Totally. I, and I think, you know, and I, I think we, we have to staff up, 'cause I mean, we, we went from, uh, you know, kind of like- ... call it zero customers to, you know, uh, 20, 30,000, kind of, you know, in, in six weeks.

We have to have certainly more, uh, customer support, customer success stuff, et cetera. But, you know, also just on, on engineering, um, we have to ramp up. But I do think that there's a, uh... Uh, we saw this in the 2021 cycle, right?

Where, you know, adding tons more people can be a thing that really hurts, you know, uh, the company because you can... I- it's just harder. It's really hard to manage lots of people. Not s- If you're a big enough comp- company to warrant a certain headcount, 100%, you kinda have to do it, right?

But, um, I think for us it's worked just to really, you know, grow s- grow the team slowly and intentionally. And so I think we're gonna take the same approach here, at a bit of a faster clip than we were previously.

But, um, to me that'd just be general advice to startups, is like, slowly, intentionally, as fast as you can to meet demand or whatever.

Part of what I felt like you're in a unique position to talk about, but also kind of what we went through in our, in our, uh, call, was, "I have PMF, now what?" Is, is kind of what I've been I've been saying.

And so, like, uh, I think y- you know, the first answer is hire a data scientist.

Yeah.

Because we have to sort of figure out, like, from our data, like, you, you're now sitting on, on a ton of different customers, and we don't really know the different customer segments. You're starting to get an idea of churn.

You're starting to get an idea of, like, segmentation. You already had data enrichment. One of my most interesting quotes from you from that session was that because you were selling to enterprise for so long, you had already set up all that stuff, and it just, like, wasn't useful for a more sort of developer, bottom-up centric approach.

Yeah, well, s- and particularly because for the first time in the company's history, we're, we're selling primarily to almost non-developers.

Yep.

Um, and so everything that we've ever... All the playbooks we had- ... not relevant here, basically, right? So the... And, you know, one of, uh, one of our investors I talked with, uh, earlier this week, uh, b- basically brought up a really great point, which is like, you are now a B2C company.

Yeah.

And how you operate needs to reflect that.

Which is, which is what? I, I don't know.

Which is basically, um, from an analytics perspective, like, it- you're tracking everything, right? And, and then to your point, you have, you have people kind of round the clock slicing and dicing data to understand who are these people coming in, who are the types of people you actually want to retain versus people that you know are just gonna churn out, and that's okay 'cause they're not the actual, like, ICP that you're going for, right?

When you're building stuff for enterprise software, the bar is a lot lower. And, and to kind of, to... From the conversation before, one of the biggest, and this is kind of what we found with Stackblitz which is kinda interesting, you know, you mentioned it, it's like, it's, as a startup, it's very hard to sell on-prem.

Extremely true. But if you can do it, it's like the promised land because, you know, these, these companies, uh, you know, they're Fortune 500s. They can write really large checks. And so when you're going and selling to them, it doesn't matter so much.

Like, on your website, sure, you wanna track the conversion to the enterprise contact form or whatever, right? But what, what actually really matters is, like, y- the, a lot of human touch points of, "Hey, we wanna have a quarterly call after.

Just getting them to install this stuff," there's a whole playbook for that. You need to hire sales engineers that can be on the ground floor and helping people install it. Then after that, you gotta, "Okay, how do we make sure they're kinda constantly successful?"

Because you can't access... Like, we can... Our enterprise customer instances, we have no idea how often they're using them. Why? 'Cause the, the whole point is that we can't see what they're up to for, uh, you know, good reason, right?

Like, they, they need to own their data. And so the way... It's actually much, a very complicated problem of how do you have, like, build relationships where you, where everyone's getting on calls, they can share kind of the telemetry that, that they can see within their instance, and you can kind of extrapolate that and make sure they're happy and successful.

So, so that's, there's a whole art of that, of doing enterprise well that, uh, we've gone and done and, and, and closed these folks. Totally unrelated to doing BC. Completely, uh, completely unrelated for the most part. So anyways, that, so that, you know, we're, as a company, we're, we're kind of reorienting, you know, our focus on, uh, okay, going and actually really leaning in on analytics, whatever have you.

And fortunately, like, my co-founder and I, the, our, the enterprise business of Stackblitz was the first time we had ever done enterprise. Primarily, like, things through the company we did before was B2C. Like, we were selling people courses on how to do web development, basically, right?

So a lot of the skill set that, uh, you know, I had built up there, I able to pull that back off the shelf, dust it off, sharpen the blade and you know, say, "I'm, we're doing email marketing.

We're doing live streams," you know? So anyways, so that's, it's, it's kinda cool to, you know, uh, be shifting back to some of the, the stuff where we cut our teeth on back in the day.

Wow.

How did you pick the pricing? Because I had to pay, uh- And, and, uh, and there's like-

You know, that's fantastic. That's fan- You wanna, like, slight- slightly, like, uh, uh-

Yeah.

It's like you gotta pay. It's like-

Do I have to? Okay

... it's like you're running out of tokens, dude. I was like, "Fine, I'm running out of tokens?" It's like I don't wanna run out of tokens. But there's like five different tiers.

Yeah.

Right? Which are kinda like token based and capacity based.

Yep.

How do you kinda reconcile that in the consumer side, where maybe the consumer doesn't even really need to know what a token is, right?

Yeah.

Like, on a, like your mom probably doesn't really care-

Nope

... what a AI token is.

Yeah.

How did you structure it to start? How did you come up with that? And then maybe ideas that you have to, like, improve or, like, modify that.

Totally. Yeah, so we, so when we first launched with Stackblitz is, like, g- we were an enterprise play, right? And so, uh, when we launched in 2017, I think we tried pricing in 2018 or 2019.

Mm-hmm.

Like, it was free for a long time, and then we had a $9 plan, and that was just the way it was. It was, it was kind of like our, our $1.50 hotdog at Costco.

Right, yeah.

It's kinda like this, this, you know, just low price, just, you know... It, it wasn't the primary rev driver, and we just wanted, you know, to say, "Hey, pay for some more storage and private projects or whatever." As we went to launch Bolt, again, like, our expectation was, "Hey, we're, we'll probably get a good number of people that'll sign up and be excited about it.

Um, you know, we're not too concerned." You know, we're just, we're just not... We were unprepared for the tsunami that hit. And so after going online, the first week we were like, "Wow, this is cool. There's a..." I mean, it just kept growing.

And then once we hit week two, I mean, we were just... Th- nine bucks was... I mean, it's like the cheapest AI coding thing you can get, maybe other than Copilot. But, like, we were overrun by support tickets.

And it just, and just the sheer volume of people coming in, it's just laws of supply and demand. We were like, "Okay, this isn't, we- there's no way we can scale to meet this." Also, the people coming in are burning through their tokens, and there's no way to actually, like, buy more of these things.

And nine bucks is just, y- you can't get that much inference out of that. And so the- here's the other thing that, uh, that's interesting about Bolt compared to, like, something like Copilot or whatever, and this kind of tie, th- this, sorry, it's a little bit of a roundabout way to answer your question, but basically what we ended up, at that moment, we ended up realizing is that when you use Copilot, what it's sending up, it doesn't provide a lot of context of your code base.

They try and reduce the amount of context as much as they can, and I, I, I think, you know, the origins of this stuff is they, everyone kind of wants this, like, low price point-

Mm-hmm

... where it's, like, all you can eat, so it just kind of f- that kind of feels like... 'Cause it's like, it, it, almost like Netflix. It's like, "I'll pay a thing, and then I can just do as much of v- v- you know, movie watching as I want."

And I think, I think that, that kind of mentality, when these first AI products came out, kind of makes sense. They're like, "Okay, well, we, we don't wanna meter it 'cause that doesn't feel good," right? But the problem is that then they're incentivized to not have it be able to capable-

Right. Yeah, yeah

... like, the more context you give it, the more it can do. And that's the magic of, of what we're doing with Bolt, is we're giving it all the context we possibly can. And that's why you can go to it and say, "Make me an RSVP site," and it does it because it has context, the en- the entire state of the application, you know, et cetera, et cetera.

And that's what makes it so accurate versus if you go to Copilot and say that, it, it'll be, you know, it might punch out a, a React component that's the button to create the thing but not actually more than that.

So anyways, so, um, you know, and, and at the time, uh, when people had bought the $9 plan, they were like, "I want to give you more money. I want to buy more tokens. How do I do that?"

And so our team scrambled. That weekend we just turned it around and just, you know, we said, "Okay, well, what do we think is reasonable?" And, and, and we said, "Okay, so let's go, you know, immediately double the prices of the, of the base tier because it, it's just not enough-

Yeah

... what people are getting on, for nine bucks. So that'll be... That seems reasonable. It's kind of in line with everyone else." And then we added 50, 100, and $200 plans 'cause we were like, "That should be enough."

And so, yeah, so that, that's kind of the origins of it. And, and, um, it was, it was people that use it fall in love with it, and they wanna use more of it, and the problem is that inference is expensive.

And so we're not actually taking, you know, to date on the, on the revenue we've done, we have not really taken a margin at all on this stuff 'cause we're just trying to put all the value back into the folks that are there using the tool and just getting the maximum amount of value out of the...

But it's really key to the, kinda the magic of, of the experience. And so the other, the other thing kind of worth mentioning is there's kind of the ARR number, but then we, you can also buy additional tokens, you know, just the way, with the usage-based billing, effectively, and that's accounting for an additional 20, 30% of, of revenue that, that's coming to the company.

People are actually using this to do their jobs. Like, you, think, think about a web development agency before this thing. They're going and using Figma to make a design. They have to pay a designer. They have to, like, punch that out into code, kind of manage.

Maybe, like, Copilot can help a little bit with punching out this thing, that. They're coming to this thing and th- there's just wild stories online where it's like guy, uh, a bake- local bakery's like, "We need a website."

He's like, "Okay. Well, I'm gonna charge you 1,000 bucks." They're like, "Okay, that sounds great. Reasonable price." 30 minutes later he's like, "Here's a deploy preview of your thing. How's that look?" They're like, "Wow. Holy crap. This is so fast."

I'm not giving you 1,000 bucks.

But they did. They were, they were, they were-

Yeah, yeah, yeah

... mind-blown 'cause they were like, "This usually takes months."

Yeah, yeah, yeah.

You know? So some of the biggest power users are people that build websites for a living because this is the, the alpha on this is insane.

That's almost, like, the gap, right? It's like it used to be that if I ask you, before this, to do a website, and in 30 minutes you return to me and you give me something, I'm like, "You know, you're probably just copying something else you've done before," you know?

Yeah.

Versus now it's almost like it doesn't really matter how much time it takes you because everybody's gonna be so fast with these things. It's more, like, the value. And that's why in your pricing tier it was al- almost like there's only really gonna be, like, either $20 a month users or, like, $1,000 a month users.

Yeah.

You know, it's almost like who's gonna use the $50 a month because it's kinda like in between, between being infrequent user and being, like, a power user.

Yeah.

You know? So yeah, it makes sense that you have, like, a big part of, like, on-demand on top of the tiers.

On top of it. Yeah, and, and on the 50, the, the, there's actually a lot of people on the one. I think it's 'cause it's, like, enough to actually, like, for developers that are using this to just kinda, like, punch out components or designs or whatever, kinda gets them enough for, you know, kind of in a given month or whatever.

And, and so it's been interesting to s- kinda see the, the, you know, the, the, the upgrades that happen.

Yeah.

But what's been kinda cool about the product is it's, um... And again, I, I think this is kind of novel, and this is, you know, us being maybe a little more transparent than we should be or something, but, like, I suspect we're gonna s- I think we're gonna see a lot more of this because we're hitting an inflection point.

Coming back to the Copilot thing, part of the problem before is that it didn't matter if you provided more context. The models just weren't good enough to know what to even do with it. That's not the case now.

You know, uh, just one, one, uh, you know, story of, like, one of the first people, uh, one of the power, first power users, uh, that adopted Bolt was this gal in Thailand who's a PM at a software banking company, and she had an idea for this app called ViralHooks.ai, which is basically, it's a tool that, uh, if you wanna make viral TikToks and stuff, it's like what's the hook of the video to make people watch, right?

And so sh- basically, uh, she, you know, you can go and, like, see, it, it goes and, uh, extracts hooks from other people's videos and helps you with, like, you know, AI to write your own. And, um, she had originally put, the week before Bolt launched, she put that on Upwork, and, uh, you know, uh, some, uh, I think a developer in, like, uh, Ukraine had quoted her, you know, $5,000, and it was gonna take, like, three months or something like that.

Reasonable timeframe, right, for an app like that. Reasonable price. The week after that, uh, Bolt came out, she bought the $50 plan, and she had the app built within a week or two. And so you're talking about, like-

This is Bolt

... that's it. And it's beautiful. It's, it's, and she, she did an incredible job, right? And so the numbers are, are wild. $5,000 three months to $50 and, and, like- A week.

Yeah. You gotta charge more, right?

Well It's- but so, so it's, it's kind of like, so the, you know, there's, there's people, like when- when- we've had a lot of people go, "This pricing's insane." And we're, and we're like, "Well, we're not even taking really a margin at the moment on it," you know?

But also, but when you, when you compare that to the price of actually going and building... The cost of building quality software today, anyone who knows the price of building quality software, the alpha is obvious, right? It's a 99% re- you know, cost reduction, and 5X faster, uh, you know, delivery time, you know?

So anyway, so that's, I think we're one of the first products that have actually come out kinda proving that, you know, in, in, in a revenue way. To kind of underscore the point, as you can imagine, we've had s- you know, kind of ven- you know, venture capital firms kind of reach out kinda, you know, curious to kinda, you know, what we're up to or whatever.

And so, you know, one of the most, uh, you know, there's kind of, one of the, the most notable ones or whatever reached out, and so we kind of sent them, you know, uh, you know, kind of our numbers.

Um, actually it was the investor update, Sean, that, that I think you pr- you know, the, you know, the one you saw. Kinda gave them a snapshot of it, and they, um, one of their analysts accidentally replied all-

Oh

... on what we had sent them, and with, with the analysis. And so on, on this part there you know, one, one of the things they said was, "We haven't seen anything that's kind of eye-opening to see people going to $200 tier on this sort of thing.

Haven't seen anything else like that in the space." 'Cause I, I think this is very new because of the new model capabilities, right? Where people, you know, it makes sense, like you're willing to pay more money for this stuff, so.

Yeah. This is something I've talked about before in terms of matching the dollar amount of spend to the capabilities of the AIs. The chart that I published in, in the past was, you know, OpenAI has like five levels of AGI-ness, and level, level one is sort of like a chatbot.

Level two is reasoning. Level three is agents. Four is organizations. Five is some- something super, superhuman. I don't remember what the, the exact levels are. But each, each, you can sort of each match each of them with like tiers.

Like, uh, $20 is like the ChatGPT tier. $200 is where you're at. $2,000 is higher. 20,000, 200,000, right?

Yep.

Like the, you can see levels where it makes sense. Like, you know, I think Brightwave is also there, by the way. Like I, I don't know what Brightwave charges, but it's higher, right, than a ChatGPT, and like you have to deliver more value for that.

Yep.

But you, you can do it now.

Yep.

So then why not? Maybe everyone should do it.

I, I think we're gonna see a lot more of this. I think we're gonna see, and I, I, I think, you know, for AI codegen specifically, this is the first moment where I think that there's been that moment where, where it goes from zero to one, where it's like, yep, the price po- you know, the value, the value is so, like of what you can get out of these things, is so much higher than it was, you know, three, six months ago, that, uh, I, I think we're gonna see, I think we're gonna see a lot more of this.

Like we might, we, you know, s- you know, Bolt is, is I think one of the first things. But, uh, yeah, I mean, it's just, uh, to me it's, it's, it, it's inevitable that we're gonna see many more things kind of leveraging this, this sort of use case and the amount of efficiency you can get out of using these systems, right?

So-

Yeah. Yeah. Yeah, because, I mean, the Bolt arbitrage would be, quote, the price based on the query, you know? It's like-

Yeah, you're saying high-value tokens.

Yeah.

Yeah.

It's like, "Hey, um..." It's like your mom, it's like you wouldn't charge your mom $2,000 to tell her stories. But like, you know, this person doing an app and like a product on it-

Yeah. I mean-

... you gotta pay more, you know? But it, it's hard right now. I understand. It's like, it's really hard to figure out how much you can push it, how much value the person will get out of the thing, you know?

Yeah. So I wanna riff a little bit on, like stuff like this, right? I think you nailed a lot with the design system. You know, one of the differences between open source Bolt and the one that you have is actually, like you, you spent a lot of time on, on, on the design system, I think, right?

Most things just look great when they come out, but then there's also a whole backend portion that they need. Was that a challenge? Is there anything that you're sort of like figuring out that you wanna riff on? Yeah, t- so I, I think one of the, one of the main things, I think you, you hit the nail on the head, which is, you know, and kinda going into putting Bolt online, we originally, again, we, we've been selling to developers, and so we were kinda like, "This is a tool for prototyping," and they'll download their code.

And, but what ended up, we ended up finding in the early user testing was how important the deployment story was, and how... And, and, and this is something you said to me specifically. You were like, "Backend. This needs to, like backend needs to be part of this, like logging in, like auth."

I just could, to triple confirm, you were dead right. That's been, that has been the absolute number one thing that, uh, folks coming to Bolt, you know, are looking to do, is build a real app with a backend with billing.

And so one of, this guy, um, uh, Mauricio, he's one of our power users, and he's like, "There's three things that like every app that I'll ever wanna build in Bolt, and any of these other people in this community," you know?

Three things: a database, auth, and payments. Those three things, right? And so, and so that's, um-

Admin dashboard

... ad- We, we can do that pretty, pretty decently, um, pretty decently. Um-

As in e- every database needs an WP-

Oh, yeah

... WP admin.

Yes, yes. Correct. Totally, totally. And so yeah, so today I think like, like Viral Hooks, uh, for example, I think she's using Firebase for auth and database and that sort of thing. You know, so I, I think Firebase and Supabase, those are the two things that, that just work incredibly well.

And so that's actually the point where, where we're at now, where, you know, right now it's, uh, you know, if folks have to still, you know, kind of go to Supabase, manually spin up a thing, come back to Bolt, put the thing.

You know, it's like that sort of processy thing with Firebase. Um, each of those products kind of have their own little quirks, so you have to, there's like kinda steps, right? And so-

BoltBase

... yeah. Bolt, BoltBase. Yeah. I, I think, yeah, I think, um- ... initially we're like, "Okay, there should just be a way to like, uh, for Bolt to just go and spin up these things on their behalf," and just, and just, you know-

Yeah

... um, both of them have APIs to do so.

I'll go even further. Like have like, uh, pre-warm instances that you just assign. Like so it's already spun up, right? And so it's, so it's like a kinda serverless feeling, even though it's like not really.

Yeah.

But like, uh, yeah, just pre-warm and then just kind of assign it when, whenever someone like spins up an app.

That's a really great point. Yeah, just keep, keep one Firebase in the hopper, basically. Yeah.

One, 10, 100.

Yeah.

I don't know. More generally, this is what I felt that I wanted to do on our call, which is like when you have PMF, yes, you wanna invest some time in like understanding your customers and do data analytics and like tighten, tighten things up in general, like-

Yeah

... tighten up the pricing, tighten up the, the cost and all that. But then, like you also have to work on like what is next Like the next level in growth. Like, you can still infect.

Yeah.

I don't know what that is. Uh, but, you know, I, I went to... I wanted to keep pushing you, and I don't know if I, I, I did. Mostly 'cause I was, I was h- h- um, serving as facilitator on that call.

That's what I think. Like, I, I think you gotta st- still keep pushing the frontier-

Yeah

... and I don't know what it, what it, what it is. But, like, you know, I wanna hear what you got-

Yeah

... thinking about.

I, I think there's... You- s- we've addressed just a lot of the low-hanging P0 stuff, and, and, and we've actually seen, we've kind of the, the, the... Yeah, there's, there's key moments-

That's gonna be huge

... where it's just kinda, like-

Yeah

... been going like that, which has been cool, 'cause it's like, okay, well, the- we- we're, we're just getting started. This is just the, this is just the fixing obvious things part. Fundamentally, I think a lot... what a lot of people are coming here to do is just how can we just make it faster to go from idea to production, and a lot of it is, like, uh, I ha- when I have to go to Firebase, Supabase, spin something up, run a migrate, you know, like add a table.

The, it's like, the agent can do that, you know? So that stuff should be baked in. You know, same thing with the, the deployment side. It's like right now it's going to Netlify, but you have to create a Netlify account and go and, and do that, right?

And so I think one of the things we're gonna end up doing here is just having the hosting, um, be baked in. And so I've, I've, yeah, I've been talking with, um, um, with Matt over at Netlify about this, 'cause they actually have a way to kinda, you know-

Yeah, the white label stuff, yeah.

Yep, the white label stuff. And so... 'Cause people are- they're just trying to make a website, you know? And so it's, uh-

Y- I mean, that means also you take over domain registration.

Can you imagine, right? Like- ... you know, a couple months from now you come to this thing like, "I wanna make, uh, I wanna make an RSVP site," right? And it's like, "Uh, great. Do you, you know, do you have a name for it?

Do you want a, you know, a domain?"

Yeah, yeah.

You're like, "I, I don't know a name." It's like, "Well, here's like 10 options, and the dot coms are available. Any look good?" "Yep, that one does." "Okay. Do you wanna buy it?"

Yeah.

"Okay, great. It bought. The DNS is pointed at the thing. Should we start building this?" "Okay." "Does this look good?" "Yep." "Okay. Am I okay to push this to prod?" "Yep, that looks good." You know, like that's without leaving the product, right?

Um-

So, so to me, like Itamar was the first to actually say, like, "You are the new Wix." I never, I personally never thought about it that way. Wix is a $10 billion company. Where you wanna go, you know?

'Cause you still have a choice here.

From what we're hearing from the folks using the product, I think, um, uh, I, I don't even think Wix is even able to solve their need, you know?

Yeah.

Um, eh, but, but not to say that we don't wanna, you know, that, that, what you're saying is not what we want, but, but I mean, yeah, like I think we wanna solve folks' problems, and I think that there's a huge gap in the market of being able to build, you know, kinda more sophisticated high-quality software like websites in a way that w- for someone who's a non-engineer.

And so I think there's a huge market for that. And obviously even if you're trying to build a, a wedding website, yeah, this is, this is easier and faster, right? So I love it. I, you know, again, coming to the origins of why, uh, Albert, my co-founder, and I are doing this is we've always just loved building stuff on the web.

It's like this, I, this is the tool w- from w- even when Stackblitz was just the IDE interface to the technology, it's like we- this, this is the thing we wish we had when we were 13 years old, you know?

And, and with Bolt, oh my God, this is the thing I wish we had when we were 13 years old. Th- I'm so glad that my daughter's gonna have this thing, you know? So anyways, yeah, I, I think it, it, it makes me pretty, pretty stoked that, um, people are gonna be able to actually build amazing web applications.

Um, they can do really sophisticated things, you know? So yes, I think the short answer is h- heck yeah. I mean, if, you know, that sort of market, it is totally right up our alley.

One other angle that I wanted to pursue was also the other languages. Uh, you know, you're very JavaScript centric. We've talked about Python forever, Ruby maybe. Is that important? You know, like the previous generation of, uh, site builders were mostly Ruby shops-

Yeah

... and then m- some PHP. Do we want to capture that, or are we just like, you know, always been on JavaScript and just let JavaScript take over the world?

You know, I think, I think we're, we're s- we're certainly with great interest interested in like, you know, other languages, and we have like y- minimal support of, uh, Python and some C++ stuff in WebContainer, uh, that you can like run or whatever.

I think especially with the, with the stuff we're seeing, though, it's the language is, is kind of ancillary to the, to the-

Yeah

... to the thing.

Well, there's the ecosystem of like, I wanna end up with a code base that I can hire humans on to do the stuff that Bolt cannot do.

Yeah. True, and I, I think in that sense, like the, the, the JavaScript Node.js ecosystem is huge and well established. So it's I think you'll be a- certainly be able to get people to work on this stuff. And I think the only thing that would be missing is it's like, are you building web apps that, where a lot of the functionality is only in libraries that are in Python or something, right?

And I, I think just kinda seeing the, the applications that are being built here, you know, I think that'd be like data science and like-

Yeah

... ML and that sort of thing. And so that's, we're not seeing a l- a lot of that stuff, you know? And then but I think that's like worth like k- kind of a more generic approach as like what Replit's doing, where they're spinning up r- you know, real VMs.

You can kind of run anything. And I think they started off with like doing Python serves. I actually, I even tried, um, their, their, you know, their new agent stuff that's, uh, based on, um-

Replit Agent, yeah.

Yeah.

We're, we're close friends.

Yeah.

Replit has the database, the sort of live hosting, everything integrated that you're going to want to build, and you're, I think you're on a collision course with them, to be honest.

We'll see, 'cause I, I'm curious, uh, and you're not, not the first person to, to, to say that. I'm curious to see how it shakes out, 'cause I, I think the challenge is focus. You know, when you are...

Uh, what, what's kind of the end goal that you're shooting for?

Yeah, Replit's firmly for developers. You're positioning it for non-developers. Like, that, that, that's-

Yeah

... legitimate.

And even, even getting, even if ch- focusing on a language or an ecosystem as well, 'cause again, the problem is the, these things can just break in a million ways, and so it's part of the, uh, a lot of the work, uh, in making the experience better.

Like, how do you get s- like how to make it, someone get an idea into their fingertips and live on prod, right? There's so much stuff in between there, and a lot of it is just errors that happen and how do you handle those.

And a lot of that comes down to having a giant database of common errors that you can, uh, maybe even fine tune stuff on at some point, right? So doing that on, on one ecosystem, you can move a lot faster than if you're trying to support a lot of different languages.

However, it's a, to the point of if, if you're kinda targeting developers, they may not need that level of kinda streamline, you know, thing.

Hmm.

I think that's kind of where I see the main di- divergence is that we are unabashedly, uh, focused on this ecosystem of bu- for building web apps.

Got it.

Um-

Yeah, you supported Vite forever.

Yeah, and, and so I'm very curious to see how, just how it all shakes out, 'cause it's, I, I think what they're doing is actually... I mean, I'm very curious to see what Microsoft does, because if anyone is good at- Giving out VMs, tying it to a coditor and, and putting AI in it.

It's Satya. He's got a cloud, he's got VS Code. They've got Codespaces. They've- they're in OpenAI. Now they've got Anthropic in, in, uh, Copilot. I mean, I must imagine, I must imagine that they're cooking stuff over there, you know?

Well, we'll make sure to ask him. Yeah, we have many friends from Microsoft listening to the pod.

So just to wrap, I don't know, is there anything else Bolt related? I just have one personal question before we wrap the pod.

Maybe like just a- advice. Like now that you've been through this journey, right? Advice to your former self.

Oh, okay. Um, yeah, at which point?

Yeah. A- advice to yourself like thinking about, there are many founders out there with, uh, with a business where they're like, they're, they're working really hard at it. It's interesting, but it's not an AI business.

Yeah.

And you kind of took the plunge to, to invest in this, and it worked out for you. Maybe a lot of people are like, "Okay," like, you know, "This guy got lucky." Obviously, there's a little bit of luck in everything-

Yeah

... but, like, how do you improve your chances? Like, would you say go for it? Would you say everyone should go for it? How would you advise someone who was in your shoes and thinking about, you know, "Maybe I should have a second product.

Maybe I should take this, this experiment," or, "Maybe it doesn't work out." Like, what, what is, what's the calculus here?

Yeah, we were deeply skeptical going. I remember even the conversation you and I had. You know, I was like, "This, I, I think there's something here." At that point, we had built some amount, but, um, I had waited a long time to, to give you the call.

I said, "This is your moment."

Well, it was.

It was the fucking moment.

I remember specifically though, at the beginning of the conversation with Sean, um, he and I sat down at a coffee shop in, in, in SF and, and so I was kind of giving him the pitch of like, you know, "I think we have, I think the..."

I can't remember the exact framing I said, but it's, it's... It was obvious that Sean had heard a lot of people say this exact thing to him over the past year or two, which is like, "Hey, man, we've got an AI play.

Like this is our thing plus AI equals this, this could be crazy." And Sean, I- you gave me this like skeptical look and then, and I was like, "I really think so." And kind of, "Here's why," right? And, um, and I think, I think that's, it's actually, I think it's, that is a, internally having, being skeptical of, uh, just kind of going and jumping on hype trains is, is good 'cause it's like I think you, you know, your focus and your time and what you're putting your weight into is, uh, the most important thing when you're a founder.

I think for us, like we actually, again, like I had mentioned at the beginning of this, you know, we had tried Bolt and didn't see the results and that was like a two-week sprint and we rolled it back, right?

This, this isn't, you know, viable at this point. But, uh, but then when, you know, once we, once we saw real tangible results of, you know, some of the new stuff, we're like, "Okay, that, that changes things." And I think a lot of it is, is too, is going and finding that out for yourself and then going and talking to the smartest people you know with more domain knowledge on that stuff than you have and going, "Here's kind of what we found.

Does this track?" So when Sean and I met and, and he, and you know, we he- he and I kind of, he saw it, we talked through it, and he said, "This is your moment." I specifically remember that 'cause I, I walked away from that and I was like, "Holy shit, this, this is it."

Like this, you know, like Sean's, Sean's at the intersection of web and AI and has like, it, you know, has one of the best perspectives on this stuff of, of anyone I know. That put a huge, you know, wind in, in our sails honestly of just like, okay, let's, let's go and really, let's go and double down here because, um, you know, we had conviction before, but having someone who's in the space independently kind of verify meant a lot, you know?

So.

Makes me uncomfortable, but thank you.

I mean it. I mean it.

Yeah, yeah.

You know, and I, and I waited, I waited until I was pretty darn sure I was, it was not gonna be a waste of time to-

Yeah

... to put it in front of you, you know?

Cool. Well, that's all I have.

Um, yeah, and then on the personal side, you had a baby in April.

Yeah.

You ran an Ironman in October. Now it's November and you're-

He did Ironman while launching Bolt.

Yeah, exactly. That's what I'm saying, like-

I was trying, I was trying to schedule the call for him, and he goes like, "Nope, I'm, uh, I'm running."

Sorry, I'm swimming. He's like, "Hey, I'm on the swimming session." Okay.

And, and for those who don't know, uh, actually, I did not kn- I did not know the distance of an Ironman.

Yeah.

It's 13 hours. Your time was thir- 12, 12, 12:30.

Yeah, 12:15. Yeah.

12:15.

12:15, 12:15 in mine.

Okay. Oh, wow. Ooh.

You're like- Every minute matters. Give me my minutes. Give me my minutes. No, no. I, it, it can, it completely depends on, uh, you know, the course and just the, the, the person or whatever, right? And, but yeah, I mean, it's, it's, um-

So yeah, 2.4 KM open water swim.

That mile. Yeah, 2.4 miles. Yeah.

Mile open water swim. 100 KM, 100 mile, 100 KM cycle.

Yeah, it's, I think it's, I think it's 112 mile, uh, s- uh, bike.

And then marathon.

Then a full, yeah, full 26.2 mile marathon. Yeah.

Crazy.

It was-

Why?

Yeah.

And, and you were n- you were not like a super endurance a- athlete before, right? Let's, let's like make this clear.

Yeah, kind of a wild, a wild thing. So, you know, back when I, uh, we, we had our daughter, um, in April, and at that time we were, the future of the company was, you know, we were, we were figuring out what are we gonna do here.

At that time it was, it was pro- just prior to Bolt kind of getting kicked in, you know, the rebirth of it with the new models and stuff. And so I, I knew that it was gonna be... You know, having a, having a, a, a child is, you know, if you talk to anyone that's done that, you, you don't have a lot of sleep.

It's, it's, you know, there's a lot of, you know, to, to, to be a great parent is, is a ton of work, and then also being a startup CEO where there's a lot of uncertainty or whatever. The way I've always found, like when I have to go and, um, kind of knock it out of the park in all aspects of my life is, is, uh, going, yeah, just to, to make it all aspects of my life.

And so I was, I, I just one, you know, woke up one day, I was like, "All right, I'm gonna do an Ironman this year." And, uh- ... uh, burned the ships, bought the, bought... It costs 1,000 bucks to do these.

Didn't know that. Um- And, you know, just started. I, I never ran a marathon at that point, and so I think it was like thir- 45 or 60 days after that I ran a marathon. My brother-in-law, he's, that was even more insane.

Two weeks before the marathon, I was like, "Hey, you wanna run a marathon in two weeks?" He's like, "Sure," and, and just did it with me. He, not an endurance athlete either, right? But anyway, so yeah, so I, I, I was training.

Ended up getting a, a coach, uh, who's, uh, usually go, you kind of online. He's up in Marin. Great guy. Was on the US Olympic team for triathlons, and when I told him, "Okay, I'm, I'm doing Ironman California, um, in three months," he was like, "Uh,"

"Are you insane?" You know, like, "What are you..." You know, "You didn't ask for my opinion, but like I just want you to know, I don't think this is a good idea. I, I think," you know, like, "You shouldn't do this," et cetera.

And I ended up doing it. You know, I ended up getting it done. And so he was like, "Okay," like that's pretty-

But what makes you, what makes you ignore expert advice here? Like most sane people would be, would be like, "Okay, I mean, you know what you're doing," like, "Oh, maybe wait a year."

That's totally why I think, and this is, this is kind of the, and being a founder, right? It's, it's all about like if you, like I mentioned earlier, it's like when we talk to people that worked on browser engines, they're like, "You can't, you can't build what you're talking about."

I think the job of a founder is, is to, is to solicit that advice. And, and what my coach actually said, he was right about certain things. There were certain areas where I was under-indexed on. Like, I was not, you know, spending nearly enough time on my bike, for example.

Like after that, I, I was on my bike six hours a day on the weekends. That's a lot of time to spend in the saddle, just like, just kind of, you know? And that was like, you know, for a couple months leading up to it.

He was right on, on certain aspects of it. And but I kind of had to look internally and go, "Okay, like what is he kind of missing about who I am and like what I kind of know I'm capable of at this point?"

I, I mean, it was a nail-biter. I mean, going into the thing, I, you know, it's, you get... And this is the same thing with launching Bolt. It's like, or, or launching anything. You g- they, launch day, race day, you kind of go in, you're like, "All right, here we go."

Like we're gonna, we're gonna find out. We're gonna find out, you know, how based in reality I was about all the decisions that led to this moment. And so anyways, going and doing the Ironman in like six months, most people spend, uh, you know, the, the folks he trains, usually it's, you know, one to two years on this stuff before you do, try and do a full.

You know, it's like going and kind of doing it in that sort of timeframe, it's, it, it's very similar. The same sort of skill set of going and building products. You have to really kind of look at the base reality and go make your own assessment on it, right?

So.

Cool. Well, great story to wrap. Thank you so much, Eric.

Thanks.

Thanks for your time.

Heck yeah.