Code Metal raises $125M Series B to automate translation of AI code into edge-optimized hardware

to be in the TVPN Ultra Dome. But the Lambda Lightning round has begun.

Here we go.

What's happening?

Thanks for hopping on.

How you doing?

Great. Good. Great to meet you. first time in the show. Please introduce yourself and the company.

Hi, I'm uh Peter Morales. I'm the CEO of Code Metal. It's great to meet you guys. Excited to be on the show.

Extremely metal name. I'm sure you've you've heard that before, but I had to do it. I love the name. Uh yeah. What uh what got you into this? Talk about your journey, how you got here, and then we'll get into the business.

Yeah. So, um at least for me, I started my career in defense. I'd uh graduated with a physics degree and this was one of the areas that was like hey we want people with that background and uh turns out it was an awesome area to jump into. You got to apply first principles but then also build systems that had impact and so spent some time developing basically taking software that was written at a high level and bringing it down to low-level hardware systems. And the reason being was was a big programming nerd but also kind of understood the math and it was this sweet spot of translating software to hardware. So I got to work on some cool AI algorithms on the F-35 and that kind of inspired me to do more work in that space. So I went over to MIT Lincoln Laboratory and started doing research in AI. Got to be one of the founding members of their AI technology group and then ended up going to Microsoft where I was hoping to put all these pieces together and get to build like sort of more for uh commercial industry products that took sort of state-of-the-art algorithms and brought it down to hardware. And so I was working on the hollow lens and um hoping to learn kind of these like industry best practices and it was the same as defense man. It was just as like hard to push the latest and greatest down to hardware and you know inspiration behind code metal was there's got to be a better way to do this.

Makes a lot of sense. uh before we go further uh I don't know what or how much you can share around uh AI and the F-35 but generally for those types of exquisite systems like how you know back when you were actually doing that work uh where where is AI actually uh be impactful in an exquisite system like that?

Yeah. So you got to view it from like our perspective which is like hey we're going to take these at the time kind of really cool algorithms and bring these down to ancient processors. So exquisite system but whole uh hardware systems to actually get the algorithms running. And uh this was part of the DARPA arc program. So it's like adaptive radar countermeasures. So, the idea being is, hey, I've spotted you in the sky, and I'm going to go ahead and like try and take you out now, but maybe I don't know the complete parameters behind what that system's doing. And so, I need to kind of on the fly come up with a strategy to defeat um you know, those missile sites on the ground. And so, applying some different techniques uh to that I can't talk about um was sort of the the gist of the game.

Fascinating. Uh give us the quick history on Code Metal then. when when did you actually start the company? You're coming on today with some big news, but uh give us give us uh the history until now.

Yeah, so it's been a real rocket ship. We started just two years ago.

Wow.

Overnight success.

Nice.

So um yeah, we we've we've had really really great partners. J2 uh Ventures did our preede round. Cool. and um you know their roll your sleeves up kind of like help you founder friendly fund and we jumped right into it at Shield Capital who really understood the dual use mission and then you know once we started seeing this take off and sort of for us and I'll get into what we do exactly um you know in some of the mission critical industries that we operate um you know Excel jumped in and it's just been sort of you know milestone after another in terms of like you know the real the real problem that we're solving which maybe I should give some background on next.

Yeah, please. Yeah, go for it.

Yeah, there's this explosion, you know, of basically AI generated code and it's great, but you know, you can build an MVP.

Um, but I'm sure all of you have been working like with, you know, Claude or or, you know, Open AI and maybe you've asked it like, are you sure? Or like, you know, is this right?

Don't make mistakes.

Yeah. Exactly. Yeah. Yeah. Build me a break system. No mistakes. Yeah.

You say you're you're 100% sure on this. Yeah. It's like turns out to be wrong.

Take off from the aircraft carrier, you're good to go.

Yeah. Yeah. No crazy.

Not crazy.

So yeah. So, so the idea here is like we want to answer that question. So as much as we are an AI company and we're an AR verification company and that means you know inherently these systems are stochastic, right? And so you can't necessarily rely on AI to check AI. And so we're leveraging a lot of techniques both from sort of traditional test verification and formal methods to basically prove as you work your way through these hard software challenges that we have done the identical thing that you put in. And in this case we can't rely on natural language like like writing a prompt is not a good way to define how like a exquisite system needs to operate. And so we're not like a low code or no code, we're a high code, you know, shop and we focus on basically solving software problems that are really the real tall tent poles in production, namely moving from MVP to production um for some of like the biggest industries and you know defense being obviously one of those based on our backgrounds.

Very cool. Uh take us through the funding round. What happened?

Yeah, so exciting news to announce today. Uh we have announced our series B uh 125 million. Boom. Congratulations.

And tell us about the new uh president and COO. How'd you meet? What is he going to be doing? Why is that an important role to fill?

And I think that's the second big announcement for us. We've been really heads down, a bunch of us um you know, engineering nerds and uh you know, also starting to grow in business folks like Laura Shen. She joined us from National Security Council, had a background at Uber, and really was sort of like my main teammate on the selling side.

Um, but we've grown a lot since then, and we've got, you know, operational needs. We need leadership on sort of working with some of these um bigger clients that we started to talk to on the technology side. And yeah, Ryan Aay from uh CEO of Tableau actually joined us as president and COO. Um, and yeah, he's been awesome already. like we we we learned about him through our network and sort of trying to find the right person who had the right sort of curious attitude wanted to first principles come in and do something actually really new and uh yeah he's been an awesome force just experience working with some of you know your videos of the world the different partnerships that he's had at the places he's deployed you know he knows our customers and so having him on the team has just been awesome

assuming you guys execute your road map perfectly which I'm sure you will what does it do to uh hardware development timelines, things like that. Like are we is part of Code Metal's job to bring the acceleration that developers are feeling when they're building, you know, enterprise software, maybe they're building, you know, AI uh uh apps, etc. to bring that to the hardware level.

Yeah, you're nailing it. So like um a lot of these industries have had like you know they're not one generation behind in I want to use AI tools. they're like many generations behind even what you get if you're like a cloud developer. So, uh you know I I look at it as three different things. The first is like getting actual features acting like you're a vertically integrated software hardware company because all of a sudden you can push things directly to hardware and see those things uh run on whatever system. So that's sort of our our you know rapid development uh vertical. Then you've got code portability. So there are huge supply chain problems. So, you know, we've talked to, for instance, like the army and they've got, you know, hey, we wrote all this code for Nvidia GPUs. Like, it's ours. Like, we've we've developed it. Um, we maybe have supply chain issues here. Can we run on a different piece of hardware?

All of a sudden, that's going to gum up the work. So, we're going have to pay every contractor to rewrite the code. They could translate that. That solves programmability. That solves supply chain issues. That's another huge area of growth we're seeing.

And the final is like legacy systems. It's like there's been so much investment in code that Dow shall not touch and we just put a box around it and talk to it.

Being able to update those and deploy those has been another big area of growth for us.

It's awesome. Very important work. Thank you for

Code Metal is accelerating.

It's heavy metal. I love it.

Customers are accelerating. It's it's very metal.

Yes. And congratulations on the funding round.

Yeah. Congrats to the whole team.

Thanks so much for coming on the show.

At this at this rate, I'm sure you'll be back on

Yeah.

in Q2.

We'll talk to you soon.

Have fun out there.

Have a good one, Peter.

Cheers, Peter. Let me tell you about Plaid. Plaid powers the apps you use to spend, save, borrow, and invest securely bank connecting bank accounts to move money, fight fraud, and improve lending now with AI.

Up next,

we do not

We have Eric from Free Form joining in