Expo raises $45M Series B as vibe-coded mobile apps surge 84% in App Store submissions

Speaker 2: Mhmm. Well, that's very exciting. Thank you so much.

Speaker 1: Incredible progress.

Speaker 2: Breaking it down for us. Have a great rest

Speaker 1: Great your update. Congrats to the whole team.

Speaker 11: Yeah. We'll see you Thanks for having me on. See you soon.

Speaker 2: Our next guest is Charlie Cheever, the cofounder of XPo. Previously, he co founded Quora where he served as CTO and helped scale the platform to millions of users. What's up? How are you doing?

Speaker 10: Hey. I'm great. How are you guys?

Speaker 2: I'm good. Great. Welcome to the show. I I think most people will be familiar with your background, but would love to get the introduction on expo and go through the news today.

Speaker 10: Yeah. We just raised a $45,000,000 series b from Georgian. And,

Speaker 2: yeah, what's the shape of the company now? What's the plan?

Speaker 10: Yeah. We're about 62 people right now and we're all over the world. We've got great people everywhere. And the plan is just that we've got a ton of stuff to do and this will just let us do it. What's basically happened over the last year or so is that expo's sort of become the best way to make apps with AI. Mhmm. And so, it just gotten really popular. Yeah. And so, that's meant that there's just a ton of stuff to do. Yeah. And so we're we're working on it.

Speaker 2: So yeah. I mean, coding models are clearly aware of React Native and Objective C and Swift to some degree, but there's always been a disconnect between Xcode or, you know, actually delivering the product. And it feels like when you click on some vibe coded app, it's almost always in

Speaker 4: the

Speaker 2: web. And, you know, this we it feels like there's going to be a boom. We're already seeing some stats on on the App Store. But walk me through what it takes to actually deliver value in app creation that's differentiated.

Speaker 10: Yeah. So what we've been our our basic philosophy for the last, like, couple of years since the beginning has been, you know, that we want to take everything that's good about web development and bring that to

Speaker 14: mobile. Sure.

Speaker 3: So that

Speaker 7: it's just

Speaker 10: as easy and everything that's just easy and fast of

Speaker 2: Yeah.

Speaker 10: About building and distributing on the web, we bring that to mobile. And so a year and a half ago, I would have said, you know, what's great about Expo is that you can take your, you know, React web developer team and you can point them at mobile, and they can build a great mobile app that feels really, really good. Sure. And today, the pitch now becomes like, these AI models are so well trained on React and JavaScript, TypeScript, etcetera, that they're really, really good at this for this. But the other really important thing there is Export isn't just a way to build mobile apps with JavaScript. We kind of see saw that attempted in the past, and it just didn't work out that well because people would use, you know, the HTML five as a content delivery mechanism, it just wouldn't feel right. Just

Speaker 2: Phone wouldn't be good gap was one of the popular ones.

Speaker 10: Yeah. Stuff like that. And so what's different here is just that, like, Expo is really a way to make apps with TypeScript and React Native and stuff like that, but also lets you drop down really easily to Swift and Kotlin and stuff like that. So wherever you need to, you can go make that polish. Sort of in the same way that a developer making server side software might write a code base in a mix of Python and Rust or something like that, you can move in between them seamlessly. And that's a huge deal and means that the people who are making Apps at Expo are making stuff that's top tier, top of the App Store charts, that kind of stuff, because you can hit that quality bar that people expect on mobile these days.

Speaker 1: What's your view on how like maybe I won't ask you that, but I would say like, do you think Apple how do you think Apple Apple should be approaching

Speaker 2: The vibe 30 boom.

Speaker 1: This explosion of of new software applications. Like, they they are I I there's been so much frustration with them. They've been under pressure because of, you know, billing and and payment policies. But they certainly specifically just looking at like if you just if you 10x or 100x or maybe someday 1000x the amount of software, that is a real challenge for them. So they're not gonna get everything perfect. But I'm curious like what One advice you'd give

Speaker 10: place where I think they've done a nice job is on the Macintosh where or Mac, I guess they call it these days, where like, if you wanna get software on the Mac, you can go to the App Store on the Mac or you can get it from the web. Or you can just, know, a lot of places, like if I install Slack or Notion, I typically go to their website and download an installer and it works. And what they've they've done a really nice job of tightening up the security model around the kernel and the operating system and things like that. And there's not really any significant security problems I'm aware of on the Mac. I think that should be a really good model. And also, nobody's on the developer side, I don't know of a lot of developers who are super frustrated with distributing software on the Mac either. So I think like, from my perspective, I feel like they've actually figured out a pretty good model and they just should apply that to iOS. And, you know, it's true. Like, you know, I was reading an article the other day that I think App Store submissions are up sort of 84% this quarter. And obviously, like, a lot of that is new Expo apps. A lot of those are vibe coded. There's a lot of new app building tools out there. A lot of those are built with Expo, most of them. And so, like, yeah, a lot of people are frustrated that, you know, they're they're getting rejected for strange reasons or or they don't know what what's going on or why. And But I think overall this will get sorted out. I mean, like, there's just it's too possible and there's too much demand for building new software and building customized software and building software with bigger surface areas. Like, it's it's clearly going to happen. I don't know. I feel like sort of like when Uber launched, there was a bunch of regulatory things and policy things to find out. But, like, it was just such a good idea and just needed to happen. And so, like, eventually the dam kind of I don't if it burst, but it, like, things got sorted out and and people take Uber's now. I think people will able to make software and distribute it in the future.

Speaker 1: Yeah. What what percentage of your growth do you think is, like, agent led as in, you know, various coding agents effectively choosing expo or helping guide the team or entrepreneur to using the product? Yeah.

Speaker 10: I think it's still word-of-mouth. Like, When people make a decision big enough to sort of pick your stack that you're going to build your mobile app on, they're probably going be stuck with for a bunch of years. And for a lot of companies that decides their business, they're not going to use one thing to decide that. So I think that it's a combination of word-of-mouth, is the strongest thing. Story that we hear over and over again is sort of like, oh, our company needed to build a new mobile app because we had this older thing that just wasn't cutting it anymore. Someone on our team came in on a Monday morning with a prototype that already had six screens built and felt really great and the CEO was really impressed. And so we just kept going with that. But a

Speaker 4: lot

Speaker 10: of times that person found out about it because an AI told them about Expo or they just asked an AI, hey, how can I build a mobile app really quickly? And it directs them to it. Or they also find out about, like, YouTube is a huge referrer for us. Like, there's a lot of great content out there.

Speaker 2: Yeah. Is this is the story about Facebook writing a compiler from PHP to C plus plus Apocryphal, or is that roughly correct?

Speaker 10: I think that I believe that's correct. This guy, Heping Zhao, made something called, like, the HP HP Compiler. And they're actually kind of doing something like, one thing that's really relevant to Expo is for React Native, they've they've built this custom JavaScript virtual machine called Hermes. Sure. And they're actually working on something called static Hermes, which will come ahead of time compile TypeScript and JavaScript to native code. So, you'll able to take any React Native app. Assuming this project completes, it hasn't yet, but hopefully it will soon, you'll able to take any part of your JavaScript driven app and convert it into native code seamlessly and have that if you want to speed up that part at the cost of a, you know, more brittle, bigger binary, you'll be able to choose that.

Speaker 2: Yeah. So that yeah. That that's exactly where We're I was we're moving up these, like, levels of abstraction. You you you say that, you know, you think people will be able to build apps and develop software in the future. Do you have a view on sort of like real time UI instantiating things at runtime? And Yeah. Whether or not Apple embraces that, it feels like that's something that could be coming with artifacts that are generated on the fly based on a query. Doesn't make sense for every app, but does that seem like something that might happen in the future? I

Speaker 10: mean, basically, every big app that you know, like, does something like this. Even if you just think about something that you think of as kind of basic like Yelp. At first, it's like, oh, this is a really structured, every restaurant page looks the same. But then, they start to add more features and there's like, oh, well, this restaurant wants to post its hours in this way or there's a chain here, so we need to link to other locations. Sure. And they start to make some sort of JSON format that describes the complexity of the UI, and then they keep extending and extending and extending. And so there's an old saying that sort of like every sufficiently complicated C plus plus program contains a poorly implemented half baked version of Lisp in it or something And like

Speaker 8: I think

Speaker 10: every mobile app that gets sufficiently big basically ends up with some sort of custom dynamic on the fly rendering system for, you know, the as as their content becomes richer and richer. And I think, like, what's nice about using JavaScript and React and React Native is it's like it's a fully featured, like, you know, real programming language. Takes care of all the problems. Like, makes sure things are, you know, correctly updated, etcetera, etcetera. And like you you why not use the real thing once you're going in that direction?

Speaker 2: Yeah. What does it take to get a job with you these days? If you're a software engineer, there's lot of people that are stopping they're not studying computer science anymore. There's a bit of

Speaker 10: a wreck Yeah.

Speaker 2: And then there's a question about where young people should go. How do you counsel someone who's in college right now and interested in I personally

Speaker 10: there's three things that I look for. Yeah. I would say, like, taste is probably the most important thing. I mean, and, like, I've I've said this for years, but now with, like, AI, this becomes even more important. We're just, like, we still need somebody to say like, this is good and like, this is what we should be doing and what you're actually trying to build. And that's really hard to replace. It's rare to have someone who's really, really good to taste. And so someone who just has good judgment, that's incredibly important. Then the second thing is like we call it like high APM. APM is like a term from like real time strategy video games where it's basically like if you ever played like StarCraft back in the day or something like that, part of it is a strategy game, but part of it is just like if you just do more stuff faster than other people, you're gonna beat them. And so like we look for people that just like do a bunch of stuff really quickly but like with a precision and that you're not just doing like totally random stuff, but you're doing like effective stuff. And then the third thing we talk about is just high agency where like, you you can see right now that like writing code has become not a big problem for a lot of companies. Yeah. But shipping good software is still a big problem for

Speaker 2: a lot of companies.

Speaker 10: Yeah. And that's because the problem of like actually making stuff and getting over the finish line out the door and to your users and properly messaged and all these other things and supported and, you know, iterated on to match what your customers actually need, etcetera, etcetera, etcetera. Those are all pretty hard and a lot of times they involve like doing annoying stuff or solving like problems nobody ever wrote a manual on how to solve before. And just so people were just like, oh, I'm gonna get this done and I'm just gonna figure it out. And I'm just gonna like, you know, run through walls to to figure that out. That's that's always really important I think. And it, you know, whether it be, you know, a coding engineering context or any other role of the company, think that's super important.

Speaker 2: Yeah. It is. It's it's great. I like it a lot. Jordan, anything else?

Speaker 1: No. This is great.

Speaker 2: Yeah. Thank you so much and congratulations on the round and the progress. Very excited. Have a great day.

Speaker 1: Thanks. Great to meet Charlie. Cheers.

Speaker 2: We'll talk to you soon. Our next guest is Victor from Slash.