Adrenum raises $18M Series A to build scalable sonar hardware for undersea domain awareness

Speaker 1: Thanks, guys. We'll see if we die or not. I hope you're right. Yeah. See you back here in seventy years. Goodbye. Cheers, Brent. Let me tell you about Shopify. Shopify is the commerce platform that grows with your business that lets you sell in seconds online, in store, on mobile, on social, on marketplaces, and now with AI agents. And our next guest is Mate from Adriam. How are you doing? Welcome back.

Speaker 11: Hey. Good. Thanks for having me on. How are you guys doing? We're doing well. Great to see you. We met at

Speaker 2: a long journey dinner

Speaker 3: forever ago at this point. I think you've been on the show. We did.

Speaker 2: We also met at the the ramp dinner. I think I was the infamous founder that snapped that phone if you guys remember that. Right. Oh, missed that. This was this was told like the story. I'm not gonna say the guy's name because because It's funny. He's a friend and and it's funny and it makes him look really silly. But this this guy is sitting there. He's got the new iPhone Air. He's he's telling everyone in the room, you cannot possibly break this phone. This phone is indestructible. Anybody out anybody here, try to break my phone. Try to break my phone. This guy goes, okay. Just bends down, just snaps it in his hand. And the guy's the guy's standing there just being, like, just in shock and disbelief. I thought That's really funny. He and he was saying he was saying, I'm gonna I'm gonna email Tim Cook about this stuff. I'm I'm pretty sure that Apple never promised me. They never ran an app. It was like, it cannot be broken. Broken by a guy from the Gundot.

Speaker 1: So so so you are in the Gundot? Tell us what you're building.

Speaker 11: Yeah. We're, we're in the Greater South Bay Area. I guess you could consider it the Gundo. I mean, we're super excited to announce our series a. Yes. But How much did you raise? We raised $18,000,000.

Speaker 3: Hey.

Speaker 11: Yeah. And Zerunim builds us scalable sensing infrastructure for undersea domain awareness. So think about, like, hunt for Red October, finding submarines, unmanned underwater vehicles, semisubmersibles,

Speaker 2: and so on. So yeah. It's suited to some of systems. I'm gonna maybe put it too plainly, but we don't really know much of what's going on out in the maritime domain. Is that correct?

Speaker 11: Yeah. I mean, the maritime domain is getting more and more contested, like, daily. Right? I think we have a good understanding of what happens on the surface of the oceans. There's cameras, surface boats, but everything's kind of shifting into the undersea quite quickly. You know? It's one of the most, unexploited gaps that we have and a lot of near peer adversaries and terrorist organizations and so on and so forth are starting to move under the waves. So, like, this entire push of, you know, The US, for example, funding unmanned underwater vehicles, our adversaries are doing the same thing. And so the big question is like, do we find those things intelligently quickly give ourselves awareness of what happens around, you know, various parts of the ocean as well as like our strategic bases and so on.

Speaker 2: Makes a lot of sense. Talk about, traction since the last time we chatted. What kind of progress have you guys made? Obviously, the new funding. I'm sure you have some great momentum, but, what's been, what what have you been actually executing on?

Speaker 11: Oh my goodness. So many things. So I think from the the hardware and software side, like, I mean, we've gone through three generations of hardware in the since we last talked. We've been moving at breakneck speeds here at Andronum. We fielded it across, like, numerous exercises. Most recently, we just got back from a three and a half week deployment with, UUV group up in Keefort, Washington. So they're all the guys that run and, like, dictate how unmanned underwater vehicles run. So we were up there alongside with a few other folks, no mission ending failures for a three and a half week deployment. Obviously, constantly testing here off the coast of California, building up our repository of sonar signatures. And we do have some things that are gonna be coming up quite soon that we're super excited to announce when they hit. So stay tuned. What does integration with other

Speaker 1: defense tech companies look like now or in the future? Like, you know, there there's been like, Andoril has a platform. Palantir has a platform. I'm sure a lot of different units have bespoke systems for command and control and all sorts of different things. Does that come for free? Is that something that people are just like vibe coding integrations? Or do you have to go and, like, do a deal and and actually figure out how you integrate? Or or do you ultimately wanna be, you know, your own platform that, you know, runs as, the operating system?

Speaker 11: Yeah. I think it vastly depends on the type of technology that you're building. Right? But specifically for us, like, sonar information is so classified and compartmentalized and, it's, it's very tricky to get access on data that has been compartmentalized such as this because you have numerous third party integrators that also own that data itself as well. And so that is really the reason why we went out and built Pearl, which is our passive sonar buoy, right? As well as Obsidian, which is kind of like, you could call it a C2, but really it is like the platform that ingests all this information. There's a bunch of like super sophisticated machine learning, and effectively putting dots on a map for end users. But, the c two in itself, like, there's various abstractions of it. You know, you have the god c twos that kind of live on top. You mentioned those companies' names. Yeah. We really are trying to own the undersea domain. By building the hardware, we're building up this massive repository of sonar signatures, and that'll give us the option to then give that data to those other c two platforms if necessary. Right? So it really is like a a fine, like, a fine balance of what we have to be doing with regards to c two. But ultimately, like, every defense tech company that's out there needs to play well together because we all have to provide an integrated solution to the end user. Right? And we do have those relationships with primes and with other kind of startups as well. Yeah. You mentioned that the data is hard to get generally.

Speaker 1: You're doing machine learning, which is obviously very data heavy. Yep. Is there, like, a scale AI, like a Merkor for this type of data yet? Are you, like, buying data from, like, fishing boats? Or, like is there some world where there is a scaled data benefit even not in a in a military context, but maybe you just want to get a bunch of data on how sonar performs in a swimming pool or a lake or something that's more controlled, and then you can transfer learn that into, like, the actual, viable domain in the final application.

Speaker 11: Yeah. Sonar is super, super challenging in the fact that you can't necessarily translate environmentals to where you are deployed. Right? So like with radar, for instance, you just have air and it's pretty linear. Like the ocean does a ton of weird things with regards to performance of sonar. And so you really have to be deployed and you have to understand kind of the environmental factors such that you can perform this highly complex math that goes behind all of it. That's really the reason why we did build Pearl, right? Like that delivers that information in real time. Like we kind of get this information in sub three seconds worth of latency and we process it and we put dots on a map for end users, we have that environmental feature kind of built into Perl itself. And you can do base level extrapolation and kind of across the board with a lot of that, but it's not super, super applicable everywhere. And so, yes, there is data that we have accessed, that is more off shelf for instance, like how do whales sound. Right? Sure. Various different types of whale things. And at least in our, you know, kind of hot dog, no hot dog paradigm, we can say, this is a whale. This is not, you know Yeah. A submarine or some something like that. So it definitely adds a little bit more kind of feature rich, environment for the end user. But most of it is really hinged on the hardware that we're developing and building. Right? So in this world of AI, you can't just access a massive repository of data and do a ton of AI on it. You have to build it yourself, which is really, like, the reason why a low cost scalable hardware solution is super, super important for us to deliver that data in real time. What does

Speaker 1: actually deploying a buoy look like? Do you have to call the government before you do that? Because I I love the idea of moving fast and doing r and d and just dropping something in the water, and that's probably fine. But at the same time, like, you're you could accidentally be spying on, like, a US submarine, and they might not want you to be spying on them even if you're, like, a friendly. Right? So, like, what is the flywheel to get the permission to, you know, do the test that you need to do before you get the contract to prove that you can do it? Like, what's the sequence of events?

Speaker 11: Yeah. I mean, we're working hand in hand with a with a lot of these folks at the government. Right? And so, like, I I detailed the the deployment in Keyport, Washington. That is something that we scheduled prior going up there. You know, we do a lot of verification and validation testing here off the coast of California, but it's kind of just like put it in the water, see if it works, check it out, pull it out, and then ship it to the location with the end user and test alongside them. We've done that kind of across the coast. Right? So we were in San Diego at a large fleet exercise at the beginning of this year as well. And so it does require coordinating with end users. Right? But that's ultimately what we're looking for anyway. Like, what are the feedbacks that these folks give to us such that we can iterate much more quickly? We could do that in our own engineering vacuum or we could just be talking to these people directly, and that's what we really wanna strive to do. And we've gotten really positive feedback from them. Right? Like, other people build buoys. They're very large. It takes a lot of effort to put these things out like gigantic research vessels, you know, barges. It takes a week to pre deploy. We just kick ours off the side of the boat. Right? And so when we give these things, these buoys to the end operators, we say break them, test them, go deploy them yourself, give us the feedback such that we can put that into our loop of iteration.

Speaker 1: And they do that. They're super happy to break things. Right? And I think that's the most valuable thing that we can ask for. I love it. Well, congratulations, and thank you so the We'll talk to you soon. Yeah. Great to see you. Thank you so much. Keep snapping on iPhone. I was gonna say don't snap anymore iPhone errors, but No. Keep it going. Keep it going. If anyone has an I wants an iPhone snapped.

Speaker 2: That's your guy. They're saying Who wants that?