Full Anduril R&D Tour: Matt Grimm, Co-Founder & COO

Matt Grimm, Co-Founder & COO of Anduril, walks us through Anduril’s headquarters in Costa Mesa—a rare, and possibly first look, inside the physical systems that turn ideas into fielded products. We tour Building A, a 200,000 sq ft R&D facility with machine shop, composites, metrology, electronic warfare test environments, and a dev-test area designed to “break things” so the team can iterate faster. Grimm explains how Anduril scaled from a ~5,000 sq ft former lost-luggage garage (with mold and no bathroom) to a global footprint of ~7,000 employees and roughly 34–35 offices, while ramping production and operations. We also get hands-on with key systems and platforms, including the Ghost Shark (XL-AUV) autonomous submarine program and Anduril’s distributed compute and command-and-control stack: Menace & Titan (in partnership with Palantir). As COO, Grimm also lays out how the co-founders divide responsibilities at scale: Palmer Luckey drives new product concepts & CONOPS while serving as Anduril’s most visible public figure; Brian Schimpf, as CEO, sets strategy & evaluates global defense programs & timing; Trae Stephens leads investor relations & owns marketing and design (Ohio State, NASCAR); & Matt Grimm runs the operational core, from facilities & security to supply chain, production, logistics, & sustainment. This is Part 2 of 4 of our interview series with Anduril’s Founders, in addition to our compiled video, check them all out! Matt Grimm: https://x.com/mttgrmm Molly O’Shea: ⁠https://x.com/MollySOShea⁠ https://x.com/MollySOShea Sourcery: ⁠⁠https://x.com/sourceryy⁠ https://x.com/sourceryy 𝐄𝐏𝐈𝐒𝐎𝐃𝐄 𝐋𝐈𝐍𝐊𝐒 YouTube: https://youtu.be/7rvStMD9WFU 𝐒𝐏𝐎𝐍𝐒𝐎𝐑𝐒 • Brex https://brex.com/sourcery—The modern finance platform, combining the world’s smartest corporate card with integrated expense management, banking, bill pay, & travel. https://brex.com/sourcery • Turing https://turing.com/sourcery—Turing delivers top-tier talent, data, and tools to help AI labs improve model performance—and enables enterprises to turn those models into powerful, production-ready systems. https://turing.com/sourcery • Carta https://carta.com/sourcery—Carta connects founders, investors, and limited partners through software purpose-built for private capital. Trusted by 65,000+ companies in 160+ countries, Carta’s platform of software & services lays the groundwork so you can build, invest, and scale with confidence. https://carta.com/sourcery/?utm_medium=newsletter&utm_source=sourcery&utm_campaign=20250923-amer-carta_sourcery_data_insights https://carta.com/sourcery/?utm_medium=newsletter&utm_source=sourcery&utm_campaign=20250923-amer-carta_sourcery_data_insights • Public http://public.com/sourcery–Investing platform Public just launched Generated Assets https://public.com/generated-asset, which lets you turn any idea into an investable index with AI. With Generated Assets, you can build, backtest, refine, and invest in any thesis with AI. Gone are the days of one-size-fits-all ETFs. https://public.com/sourcery https://public.com/sourcery Follow Sourcery for the latest updates! https://www.sourcery.vc/ Disclosure Paid Endorsement. Brokerage services by Open to the Public Investing Inc, member FINRA & SIPC. Advisory services by Public Advisors LLC, SEC-registered adviser. Crypto trading provided by Zero Hash LLC, licensed by the NYSDFS. Generated Assets is an interactive analysis tool by Public Advisors. Output is for informational purposes only and is not an investment recommendation or advice. See disclosures at public.com/disclosures/ga http://public.com/disclosures/ga. Matched funds must remain in your account for at least 5 years. Match rate and other terms are subject to change at any time.

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Speaker 00

  All of those functions report into me and I kind of oversee and drive all of that. Kind of the guts that makes things happen.

• 0:04Molly O'Shea?

  You make it happen.

• 0:06Speaker 00

  That's my job.

• 0:11Speaker 00

  It's a very enclosed chamber where you're not getting waves in or out. Fury's my favorite, but this one, this one's pretty fun too. Do you feel like you're driving a very expensive forward deployed AI compute node?

• 0:22Molly O'Shea?

  I do.

• 0:23Speaker 00

  For distributed command and control of a AI-enabled battlefield?

• 0:26Molly O'Shea?

  I feel like I'm about to go save America. That's what I feel.

• 0:29Speaker 00

  There you go. You want to play the National Anthem or anything here?

• 0:32Molly O'Shea?

  This is what we have here?

• 0:33Speaker 00

  Yeah, yeah, this is, this is Venice.

• 0:36Molly O'Shea?

  Okay, but Fury is your favorite?

• 0:37Speaker 00

  My personal favorite.

• 0:38Molly O'Shea?

  Why?

• 0:49Molly O'Shea?

  Welcome to Sorcery.

• 0:51Speaker 00

  Perfect.

• 0:52Molly O'Shea?

  Matt Groh?

• 0:52Speaker 00

  Yes.

• 0:53Molly O'Shea?

  Thanks for having me here. Where are we?

• 0:55Speaker 00

  We are here at Anduril's headquarters in Costa Mesa, California, in Orange County, just south of LA. We are here in what we call our Building A. This is our research and development building. It's a 200,000 square foot building, 50-foot clear high ceilings here. Right across the parking lot from our engineering and design hub here. So the whole idea with this building is we have a machine shop, we have a composites lab, we have a metrology lab, we have a development and test lab, all of these different environments where we can prototype things quickly, test them quickly, find the fractures, find where they break, find what parts of it wear down, and then ultimately use that to drive our iterative R&D process to get products built and shipped faster.

• 1:35Molly O'Shea?

  And what are we standing in front of right now?

• 1:37Speaker 00

  We are standing in front of, this is our XLUUV. In Australia, these are called ghost sharks. Here in America, they're just called the XLUUV. This is a fully robotic submarine, so no humans on board, no people remote controlling, fully autonomous. So you give it a mission, you program some waypoints, you program a mission set of what it's supposed to look for, where it's supposed to drop payloads, when it's supposed to surface, and you send it out on its way and it goes and performs its mission completely autonomously. So these are designed.

• 2:10Speaker 00

  At our office in Sydney, Australia, which I had the pleasure of setting up a couple of years ago when we first launched this program. And now they're currently being manufactured in our Sydney manufacturing facility that just opened. It's about 75,000 square feet and is capable of making dozens of these a year. And we'll be expanding production for them to our new facility in Kwanzaa, Rhode Island, where we'll be making both our dive LDs, our smaller submarines, and these big guys for the US market.

• 2:36Molly O'Shea?

  Rhode Island is the heart of submarines.

• 2:38Speaker 00

  It is.

• 2:39Molly O'Shea?

  It's great.

• 2:40Speaker 00

  It is.

• 2:40Molly O'Shea?

  I grew up going there.

• 2:41Speaker 00

  A lot of submarines, a lot of Navy, a lot of water. So we have a new facility just in Kwanzaa, kind of on the west side of the bay there.

• 2:49Molly O'Shea?

  So all of these are gorgeous. It's a big achievement.

• 2:54Speaker 00

  Thank you.

• 2:54Molly O'Shea?

  No, congrats. But I can't imagine Anduril has always been this clean.

• 2:59Speaker 00

  No.

• 3:00Molly O'Shea?

  So can you just talk through like the timeline and like what the process was like and the evolution of Anduril from scrappy startup to gorgeous ghost shark manufacturer?

• 3:11Speaker 00

  I still think of us as scrappy startup, but I appreciate the question. I understand where you're coming from. So we're about eight and a half years old now. So we were founded in early 2017. And our first office was a 5,000 square foot, I think it was 4,900 something, just under 5,000 square foot garage that used to be American Airlines' storage facility for lost luggage. So if you flew to John Wayne Airport and you lost your bag, American Airlines would store it in this garage and just leave it there until someone claimed it or whatever. So it had mold on the walls, it had no bathroom, we had to put our own bathroom in, it had no air conditioning, it had no ventilation, it was just a like storage shed that was where we got started.

• 3:52Speaker 00

  So we started on our first day with 10 employees, just the five co-founders of us and our first five employees, and just had a couple ideas. There's a famous picture of us brainstorming ideas of what to work on with just a bunch of ideas on the whiteboard. And that's what led to originally the Sentry Tower for border security deployments and then kind of expanded from there. So from there, we then kind of expanded within Costa Mesa, we opened our DC office not that long after, and then have been kind of off to the races since then. So we're now just about 7,000 employees, got about 34, 35 offices around the world, depends how you count a little bit of different buildings on the same campus, but thereabouts. And are really off to the races scaling, especially on the production side as we land some of these larger contracts and ramping the full-scale production and getting out

• 4:36Speaker 00

  of just the like early R and D type of phase.

• 4:39Molly O'Shea?

  How do you manage all of this globally?

• 4:41Speaker 00

  I don't sleep a lot.

• 4:44Speaker 00

  And then it's a piffy point, people kind of bring this up all the time, but like it really comes down to hiring the absolute best people you can find. And finding for us, that means folks who are entrepreneurial, folks who are incredibly mission-driven, folks who are kind of like go-get-'em, sort of like can-do type attitudes. And the question when we take on big programs like this, and I'll backtrack in a minute to talk about how the program came together, the question is never like, what can go wrong? The question is always like, okay, how do we make it go right? And there's just a mindset difference between a like negative worldview of just like, oh, this will never work, here's 10 reasons why not, to like, oh, maybe we could make it work, what would it take? Well, it would take an innovative chief engineer, it would take an innovative operations person,

• 5:28Speaker 00

  it would take an innovative production scaling person, we need a team that looks like this, how can we find those folks? So finding people who we can build the company around to have that sort of positive worldview, very ambitious, full of ideas, full of kind of that mission drive, that's really what drives Anduril at its core. And I know it sounds like some sort of cheesy business school thing, you see like embroidered on a pillow or something, but it's like, people really do drive the culture. So for us, in leadership, it's about trying to find those folks, putting them in the right role, giving them kind of the right swim lanes, and then letting them run and build their teams, and the growth is working.

• 6:04Molly O'Shea?

  So we have the opportunity to interview all the founders today. So what role is...

• 6:10Speaker 00

  One of the few days we're all together, so this is great.

• 6:12Molly O'Shea?

  It's fantastic, it's a miracle, they would say, but like, what role does each one of you have? Especially now at this scale.

• 6:19Speaker 00

  Yes. So you should ask that question to the four individually. Because it'll be phenomenally interesting to see where we agree and where we don't. So a few ideas here. So first is that Palmer is our founder, and he is the guy who has really a lot of the like innovative drive behind what drives us as a company. So new product ideas, new kind of, what is in the business is called con ops, means like, how will they use this product? How does it fit into the current frame of what warfighters are using? He's the guy who kind of comes up with a lot of these new ideas and this kind of new idea generation. He's also something of our, like, wouldn't figurehead's the wrong word. I would call him more of our, like, public persona. Like, he's out there most on podcasts

• 7:02Speaker 00

  and media, doing interviews, including over the weekend at R&D F. He did a Fox News it. Like, he's mostly like our, like, pretty vocal spokesperson who's out there. Trey, obviously, leads our investor relations, and he brings a tremendous amount of expertise there. And then, this doesn't get appreciated in public enough, but he also runs our marketing and our, our design teams internally. So he has that kind of drive behind our partnership with Ohio State, or our NASCAR partnership, or a lot of the design aesthetic that you'll see in our logos and our swag and whatnot. That all, a lot of that comes from Trey. Brian is really our, like, I would call him like our strategic genius. So he's the one who is looking at, if Palmer has an idea for a new product, or any of our engineers

• 7:46Speaker 00

  really have an idea for a new product, he's the one who kind of sees the chessboard the best on what the existing programs are, what programs the Navy or the Air Force, or what the British Army is going to want, or what the Australians are talking about, or what the Japanese military is. But he's the one who kind of sees the global chessboard of the different competing programs and, and competing companies in the space. There's a lot of companies who work in the defense sector, and he sort of sees where Andro plays into that and how we fit, and he plays, he plays that game extremely, extremely well. And he's obviously our CEO, so ultimately, we all report into him. And he's the, the big boss and decider at the end of the day, when there are conflicts and things we have to resolve. And then my part, as I said, I like to joke that I'm the chief janitor.

• 8:29Speaker 00

  A lot of the operational problems flow to me. So if that's facilities, if that's HR, if that's IT, information security, physical security, like dealing with security clearances and all that's really tricky and nuanced at our scale. But more important than all that, our supply chain, our production, our fielding, our sustainment, our maintenance plans, our logistics, like, moving things like this around the world is very complicated, very tricky. All of those functions report into me, and I kind of oversee and drive all of that. So, kind of the guts that makes things happen. It is admittedly less interesting and less compelling than, like, how do we sell a program to the British Navy? But it nonetheless kind of makes the company run in the background.

• 9:06Molly O'Shea?

  You make it happen.

• 9:08Speaker 00

  That's my job. This is an anechoic chamber. So a number of our products use all sorts of interesting radars and different communications standards, so like different antennas, different chips, all of that. And then a lot of our products get into electronic warfare, so things like jamming and spoofing. So you could imagine a world like sending conflicting guidance signals to an enemy drone so you can convince it to move off target, or doing kind of radio jamming so that you can't communicate with your drones, and that causes all sorts of problems for the enemy drones. The question is how you test that? Because we're, I don't know what airport you flew into, but we're right next to John Wayne, about three miles that way.

• 9:50Speaker 00

  So you cannot go out on the roof and just like broadcast a big signal that jams communications protocols, for obvious reasons, very illegal. So then where do you test that? Is you test that in big chambers like this, where radio waves, signals just can't escape.

• 10:05Molly O'Shea?

  It's like a Faraday cage?

• 10:06Speaker 00

  Exactly.

• 10:07Molly O'Shea?

  Got it.

• 10:07Speaker 00

  Yeah. So you can test a bunch of that stuff, and it's a very enclosed chamber where you're not getting waves in or out. So you get clean environment for writing that code.

• 10:15Molly O'Shea?

  Is that part of your meditation room?

• 10:16Speaker 00

  That is very, very quiet. It is very quiet. And the joke is, if you were in there and something bad happened, no one can hear you scream, so, you know.

• 10:23Molly O'Shea?

  It's the meditation room.

• 10:24Speaker 00

  Exactly. So in here we have our metrology lab. Metrology is a fancy term for very precise measuring of things. So when an engineer designs a part and says that we need to have this hole at exactly this location, is it exactly that location, or is it slightly off? Like, what is that tolerance with which it's okay to have parts off? If you've ever tried to assemble like an IKEA dresser, and sometimes the holes don't quite line up? Sometimes it's not a big deal, you can just like thwack it with a hammer and it works. But if you're talking about a big aircraft or a submarine, like those holes need to line up exactly. So that's, in the business, that's called tolerance and tolerance stacking. And then you need to measure all of the parts that you're making, certainly in those early

• 11:05Speaker 00

  phases of R&D, to make sure that the parts you're designing are compliant with the design standards. So here in our metrology lab, these guys are using different kind of state-of-the-art scanners, where he's scanning the part and then looking at exactly the...

• 11:21Speaker 00

  Measurements of it and then that turns into an analysis on whether this part with these grooves is exactly within tolerance of what the engineer specified it to be. So they have a whole bunch of different kind of techniques for that. They've got some very precise measuring arms that can measure with incredible precision. I will get the precision wrong if I don't ask my expert over here. So you know that if this part is designed to a certain spec or a certain tolerance that the parts are within it. But it's a fairly standard lab in the grand scheme of things for engineering companies, but nonetheless makes for very cool content. And some of these are pretty new innovative techniques using weird computer vision and triangulation techniques. This one's a pretty old standard technique using a precise robot arm.

• 12:04Speaker 00

  But they have pretty cool using a camera here with different lasers to measure exactly the part and then turning it into a 3D model. So you can imagine on one hand, the engineers over there have a 3D model that they're designing. Then they send it to get made. It gets made and then it comes back and says, okay, did that manufacturing technique work? Like we wanted it to be cast or we wanted it to be 3D printed or we wanted it to be machined out of metal. Was it precise enough? Was it within spec? So then you take the physical part and do the inverse where you measure it, make a 3D model and then kind of compare the two.

• 12:36Molly O'Shea?

  Sorcery is brought to you by Brex, the financial stack trusted by more than 30,000 companies, including one in three venture-backed startups in the US. Nearly 40% of startups fail because they run out of cash. Brex is literally built to help founders avoid that. Unlike traditional banks that let your money sit idle, shipping away at it with fees, Brex's designs help you spend smarter and move faster. Their all-in-one solution combines checking, treasury, and FDIC protection into one powerful account. You can send and receive money globally at lightning speeds. Get 20 times

• 13:11Molly O'Shea?

  The standard FDIC coverage through their partner banks and even high yield from day one. With same day and even same hour liquidity, access your funds anytime. Companies like Scale AI, DoorDash, Service Titan, Hims, Anthropic, Flexport, Robinhood and Plaid. Trust and use Brex. Start today at brex.com/sorcery. That's B R E X dot com slash sorcery. In today's high-speed business world, staying ahead means using the smartest tools possible, including the powerful capabilities of artificial intelligence. Meet Turing Intelligence. Turing builds customizable AI systems designed to solve your mission-critical challenges, no matter your industry. From expert guidance to tailored projects,

• 13:56Molly O'Shea?

  Turing helps top companies realize AI that's more capable, more adaptable and more effective. With Turing, discover how AI can accelerate your business growth. To learn more, visit turing.com/sorcery. Spelt S O U R C E R Y. That's turing.com/sorcery.

• 14:15Speaker 00

  She just asked me what all the four founders did and I, it was awkward. I couldn't come up with anything for you. We're just, we're just gonna, we'll have to, we're gonna dub it in at later.

• 14:23Speaker 03

  Can I ask you a quick question?

• 14:24Speaker 00

  Yeah.

• 14:25Speaker 03

  This is bad. To cut this in half.

• 14:27Speaker 00

  Okay.

• 14:27Speaker 03

  What should I get over there?

• 14:29Speaker 00

  Machine shop that way, back to the left.

• 14:31Molly O'Shea?

  It's very important. Pencil's here, you know?

• 14:33Speaker 00

  I told you the nonsense rolls up to me, so.

• 14:37Speaker 00

  This is what we call our dev test area. The dev test area here is a, an area that is meant to break things. It's the blunt version of it. So, these guys work on figuring out all sorts of ways to break our products. So whether that's spraying it with saltwater, as in the case of our little chamber over here. I don't know how much you've spent time on boats, but saltwater eats everything. Everything. It eats water, it eats, it eats cables, it eats metal, it eats plastic, it eats leather, it eats, eats rubber, eats everything. So when our products get deployed, either on coastal environments or on boats or in the water, in the case of submarines, you obviously need to work on a lot of that resiliency. So it's one of the test chambers they use for that.

• 15:22Speaker 00

  Drop testing, over voltage, under voltage, like all sorts of difference. Part of what we use the anechoic chamber for is different kind of resiliency to jamming techniques and all of that. In our lab in here, we've got test ovens, test chambers. So if you wanted to bake a product, freeze a product, low humidity, high humidity, run it through cycles like that. This is where you do that. We have our vibram over here. So if you wanted to take our product and then shake it and see if connectors come out, see if the different motion kind of causes things to fracture over time. Vibe tables in here and then in our corner or back over there, you can kind of see we take a couple of steps forward. Our battery chargers and dischargers.

• 16:07Speaker 00

  So if you want to cycle the battery over and over and over again, get up to high charge, low charge, high charge, low charge at different kind of power settings and really kind of drive the batteries to failure, which is obviously important for testing. We would do all of that in here. So again, in keeping with the, keeping with the philosophy, our

• 16:27Speaker 00

  design engineers in the building across the street here and then over here, prototyping, testing, development, all of that. So industry at work.

• 16:36Molly O'Shea?

  So now that I have you, I just, I figured it'd be best to ask you the hardest question first.

• 16:42Speaker 00

  Okay.

• 16:42Molly O'Shea?

  This, this is something your whole team actually wanted to address.

• 16:46Speaker 00

  Okay.

• 16:47Molly O'Shea?

  And it's a little bit confrontational.

• 16:49Speaker 00

  Okay.

• 16:51Molly O'Shea?

  Is it true that you read every single Slack message?

• 16:54Speaker 00

  Oh God, no.

• 16:55Molly O'Shea?

  No?

• 16:56Speaker 00

  No, no, no, no, no. No, no, simply way too many of them. Way too many of them. No, we have 7,000 employees at this point around the world. So we're in like something like 14 time zones or something. So they're just all day, every day, all the time. So no, I can't read everything all the time.

• 17:10Molly O'Shea?

  What takes up more time, Slack or X?

• 17:13Speaker 00

  Uh Slack by a wide margin. But the, my habit, which is potentially ineffective, is I sort by unread and then I sort by most recent.

• 17:28Molly O'Shea?

  Okay.

• 17:28Speaker 00

  So I'll just, what I'm on a like between. Meetings, walking between buildings, like have five minutes or something. I'll just sort to unread. It's not foolproof, and I have to trust a little bit of the system that if people want to get my attention, they'll just send me a direct message. But it works enough to get the pulse of what's going on. So what they're referencing is that they will see me chirp in on some random thread and some random channel on like the London channel with just a comment on something and people will be like, "How the hell do you see this?" And that's what it is. It's just looking.

• 18:01Molly O'Shea?

  Well, you're always in the flow, so you do read everything.

• 18:04Speaker 00

  Kind of, sure. We'll go with that.

• 18:07Speaker 00

  We'll go with that. Good marketing, I guess.

• 18:09Molly O'Shea?

  Where are we heading into now?

• 18:10Speaker 00

  Okay, so we are in Anduril's headquarters here in Costa Mesa, and this is what we call Building A. This is our research and development building here on campus. And we are entering our machine shop here. So I like to say that here in this machine shop, we can make one or two of anything. So we've got a pretty broad diversity of tooling and equipment in here. On one side, we have the manual shop, and you can hear in the background, those are some guys welding what looks to be the frame of one of our early Menaces over there. A bunch of manual mills, manual lathes, manual bandsaws. Think of it as like a custom metal fabrication shop. So if an engineer has an idea, they bring it over here, talk to the fabricators. The fabricators can make pretty much anything. So if you want a custom gate for your driveway, all the way through a

• 18:54Speaker 00

  command and control center for deploying around the world, these guys make one or two of anything.

• 18:59Molly O'Shea?

  What have you made for yourself?

• 19:00Speaker 00

  I've made... So the biggest thing I have made is I have tuned the axles on my son's Pinewood Derby car here in this machine shop.

• 19:09Molly O'Shea?

  Bye.

• 19:10Speaker 00

  Yeah, we're big into the Scouts. I'm an Eagle Scout myself. Both my boys are in the Scouts, and we've worked on their Pinewood Derby cars.

• 19:16Molly O'Shea?

  What speed are we getting up to now?

• 19:17Speaker 00

  Well, he got third in his whole pack.

• 19:19Molly O'Shea?

  Not fast enough.

• 19:20Speaker 00

  Not fast enough, I know. It's coming soon. So we're gonna head down this way.

• 19:25Molly O'Shea?

  Okay.

• 19:27Speaker 00

  And get to see some robots cutting metal. So on that side of the shop is our manual side of the shop. Over here is our more robotic side of the shop. So we've got our water jet over there. We've got our press break here. We've got our laser cutters here and a bunch of big CNCs cutting metal that we'll talk about. And again, just to kind of play out the point, the whole idea around here is that this is not a large scale mass manufacturing kind of facility. This is a prototyping one or two of anything kind of facility. So we can even take a look at some of the parts we have here. Like, cut in the shop here, designed in our building next door that we'll see some of a little bit later, cut in the shop here and then put together, welded.

• 20:10Speaker 00

  This is probably one, maybe two or three that we'll make here. And then, once we get the design right, there are much more cost effective places in America to make these at large scale. I don't even know what this part is. But like for that early stage prototyping, having the engineers in the building there have an idea, walk over here, get feedback from fabricators, build things and have that cycle move very quick, ultimately gets us to have products to market faster. And then, in the long run, cheaper, even if this machine shop and this kind of operation is pretty expensive on surface. So these guys here are extremely, extremely talented and can make, you know, genuinely one of anything. So we'll see a couple of the parts as we continue on the walk here.

• 20:50Speaker 00

  Yeah, so we'll make everything from molds for our composite lab that's in the next room over, for parts for our XL submarine, or parts for Fury molds, or some Fury bulkheads we'll make here. So basically, big, complicated, tricky parts, and they'll send it over here. And this thing is a absolute massive beast and can cut, you know, you can obviously see the size of the piece of metal we can fit in here. Yeah, it's pretty cool. What are you guys cutting?

• 21:17Molly O'Shea?

  Oh my god!

• 21:18Speaker 00

  Oh my god, yeah.

• 21:21Speaker 00

  Yeah, but you can kind of see and get the vibe. I don't know how many of these you've seen on your tours, but they can be pretty cool when they're working at full scale. Because the platform rotates around, it rotates this way, rotates that way. And then the head itself rotates in every direction. So you can get incredibly precise, incredibly detailed parts coming out of this machine. And again, the same philosophy of just relatively low quantity, but relatively high iteration speed. There you go. So there's a fun fact with all the scrap that I really enjoy. Don't touch it because it's disgusting. It's really greasy and disgusting. But this is just like scrap metal, scrap aluminum, that you can reforge back down into usable metal. So there are metal recycling firms around here that can just come and they'll buy just like a full bucket of scrap.

• 22:06Speaker 00

  You can kind of see what they're cutting in here if you want to. These are samples of some of the parts that the team in here makes. So everything from jet engine compressor blades, and feel free, the edges are sharp, so be careful.

• 22:21Speaker 00

  Jet engine compressor blades through

• 22:26Speaker 00

  very precise helical gears. Through, what's probably coolest to prove the point is grab this guy up here. So for our bolt product, which is a kind of quadcopter with a small explosive payload on it, the engineers, again in the building right next door, had an idea for the design. Send it over here to these guys, and then these guys program the CNCs to cut the metal in exactly the right pattern. And you can see the precision of the right pattern and where the mounting plates for the different circuit boards and different wiring harnesses go. You can see all of these holes are threaded. So they'll have an idea for iteration one, and then that becomes idea for iteration two.

• 23:10Molly O'Shea?

  Oh wow, yeah.

• 23:12Speaker 00

  And you can see that's starting to look and feel a little bit like what a drone actually looks and feels like. And then for iteration three.

• 23:21Speaker 00

  You can see the next iteration, a little bit lighter, a little bit thinner, a little bit different kind of compacted design. Now, in the long run, this is the wrong way to make these frames. So, taking a huge brick of metal.

• 23:34Molly O'Shea?

  Yeah.

• 23:34Speaker 00

  And then using a mill to like cut all of these things out is very expensive, very wasteful, takes a long, long, long time. These machines are expensive, our machinists are expensive, they're very talented. So, like, taking that time to cut all of these things out is not the most efficient way. What you should be doing is casting these. So, you create a mold, you pour metal in, cure it, then you get this out at the end. The problem is that setting up those molds and that tooling for that casting is very time-consuming and comically expensive.

• 24:01Molly O'Shea?

  Yeah.

• 24:02Speaker 00

  So, it takes many, many, many months and many hundreds of thousands of dollars to get that tooling just right.

• 24:06Molly O'Shea?

  Do you work with a range view on that?

• 24:07Speaker 00

  We don't yet. They are one of the vendors we've talked to. But this is exactly what they're trying to fix, is make that problem a lot easier. So, what we do is that for those iterative cycles where we're looking at like, okay, going from this version to this version to this version, and then we fly it. And I'm sure you've heard about our testing philosophy, testing all the time and all of that. They'll go out and they'll fly and they'll say, "Ah, this isn't quite right. This angle's not right, this bracket's not right. I want to tweak this, change that." And you can see some of the subtle differences in where the mounting is and that kind of the arms are slightly different positioned. Through that iterative phase, having the engineers there, the shop here, be able to turn quick turns on these, incredibly valuable. Versus the opposite, which is like, I have an idea, I send it to three vendors,

• 24:51Speaker 00

  I get some quotes, I send them the part, I wait a week or two, I get a part back. Just adds so much latency to that system, but doing it here with the team makes it so much faster.

• 25:00Molly O'Shea?

  Yeah.

• 25:01Speaker 00

  So, and then this one is the, basically, the near-final version of what they ended up shipping.

• 25:06Molly O'Shea?

  Is this ceramic? What is on the outside?

• 25:07Speaker 00

  It's a powder coating.

• 25:08Molly O'Shea?

  Powder coating?

• 25:08Speaker 00

  Yeah, for weather resiliency. Yeah. And then the jet engine blades are the really cool part. So, thinking of the like, because we have a product line called Venom that is basically our own jet engine blades. And you hold them, see if they're on.

• 25:25Speaker 00

  I can imagine the different pieces coming together. The air inlet for that engine that kind of fits in the front like that. So you can kind of see the different parts of the engine come together. Again, all cut here.

• 25:35Molly O'Shea?

  Wow, it's fun.

• 25:37Speaker 00

  Yeah, I have to mark them as sample only so someone doesn't inadvertently put it into a jet engine.

• 25:41Molly O'Shea?

  How many times do people take them?

• 25:43Speaker 00

  More than you'd think. The original building here was the Los Angeles Times Printing Press and Distribution Facility for their Orange County edition, so I think kind of like their local South County edition down here. On the other side of the building, it's a rail spur now defunct, where they used to roll big rail cars in with all of the big printing rolls. Yeah, it's pretty cool. And then they would roll those big rolls of paper and ink and all those supplies into this huge building here, where they would do all of the printing of things. And then they would stack the final resulting papers and kind of in those bundles old fashioned into the back of delivery trucks. Here, you can see kind of this awning that creates a little bit of a loading dock vibe. Then they would fill those delivery trucks up in their own gas station here on campus, which is pretty dope.

• 26:27Molly O'Shea?

  Oh wow, that's what that is.

• 26:28Speaker 00

  Yep. And then they would send the delivery trucks out to all of the corner stores and all the residential deliveries and all that from here. So this was, a building was here forever as print kind of died in the early 2000s. The building went into kind of redevelopment and then the global financial crisis hit, so it sat fallow for a little while. A developer picked it up and was looking to do a big kind of multi-tenant thing, so like a different company in each of these little corners. They had a deal with WeWork at some point to develop part of these buildings.

• 26:56Molly O'Shea?

  Of course.

• 26:57Speaker 00

  Of course. That all sort of fell apart and then we came in and leased the whole building. So that side of the building, that side of the campus rather, is about 450,000 square feet. This building itself is another 200,000. So about 650,000-ish square feet on this whole headquarters campus. And then we've got about 2,000, 2,100 employees here, plus support staff like security and catering and janitorial, all of that. So about, you know, in the neighborhood of 2,400-ish people here every day.

• 27:29Molly O'Shea?

  And you know everyone's location at all times.

• 27:31Speaker 00

  Yeah, and everyone's name, of course. It's incredible, honestly. I wish, I can't. So we'll come out and we'll see a couple of our products from our Menace product line here. So Menace is a distributed command and control platform. So think of it kind of like having smart compute.

• 27:50Molly O'Shea?

  Sorcery is proudly sponsored by Carta. Carta is transforming the private marketplace, connecting founders, investors, and limited partners through software purpose-built for private capital. Trusted by more than 65,000 companies in over 160 countries. Carta's platform of software and services lays the groundwork so you can build, invest, and scale with confidence. Carta's fund administration platform supports over 9,000 funds and SPVs, representing nearly $185 billion in assets under management, with tools designed to enhance the strategic impact of fund CFOs. For more information, visit carta.com/sorcery. That's C-A-R-T-A dot com slash S-O-U, R-C-E-R-Y.

• 28:35Molly O'Shea?

  Some of you may not have heard this yet, but our sponsor, Public, just launched something called Generated Assets. And it brings AI into investing in a way I've honestly never seen before. Here's how it works. You type in an idea like AI-powered supply chain companies with positive free cash flow, or defense tech companies growing revenue over 25% year over year. Public's AI then dispatches a swarm of agents that scan every single US stock, evaluates them, and instantly builds a custom index around your thesis. What really stands out is how clearly it explains why each stock is included. And before you invest, you can even backtest your idea against the S&P 500. So you're making decisions with real context, not just guessing. And beyond Generated Assets, Public lets you invest in stocks, bonds, options, crypto, all in one place.

• 29:20Molly O'Shea?

  They'll even give you an uncapped 1% match when you transfer your investments over from another platform. If you want to build a portfolio that actually reflects your thesis, visit public.com/sorcery. Paid for by Public Investing. Full disclosures in the description.

• 29:35Speaker 00

  Kind of like edge compute nodes deployed. So if you want to have AI sensors in the field, you want to have the smartest possible sort of platforms for our soldiers in the field to interact with. You need to have distributed compute along these lines. So we have very mobile versions of them here on these chassis. And then we have slightly less mobile versions of them in that sort of frame. That is what's called our Menace. And then there's a variant of all of these that's on a larger platform that's called Titan. Are we still calling it Titan? Or we change it to? Okay, so we'll call it Titan. Yeah, and we're partnered with Palantir on the Titan program for bringing their software onto our kind of hardware.

• 30:15Speaker 00

  For this joint program with the Army. So, so yep, so here on the on the Menace platform, got all sorts of antennas and communication protocols around built in. Big compute nodes in the back. You can kind of hear the hear the fans humming there. And then the whole idea is that our soldiers would be deployed out and then that gives you a lattice node and lattice compute. Wherever you are in the world. And then this also filters into the broader landscape of having soldier born compute. So like the whole notion of the the IVAS program that we acquired from the Microsoft spin out. It's now called, I don't know what they called it now SBMC, Soldier Born, something, something, I don't know. They change the acronyms every year to keep us on our toes.

• 31:00Speaker 00

  But that program was basically like, could you put a lattice node on? Deployed soldier. So whether that's a wrist mounted, head mounted, glasses mounted, kind of AR type system. And then that's of course Eagle Eye.

• 31:13Molly O'Shea?

  Yeah.

• 31:14Speaker 00

  So all of this ties together into that distributed ecosystem where compute, comms, security, all forward deployed with the soldiers. So, but come on around, drive it.

• 31:23Molly O'Shea?

  Is this your favorite?

• 31:24Speaker 00

  This one's fun to drive.

• 31:25Molly O'Shea?

  It's um.

• 31:27Speaker 00

  Fury is my favorite, but this one, this one's pretty fun too. So.

• 31:31Molly O'Shea?

  Okay.

• 31:32Speaker 00

  Do you feel like you're driving a very expensive forward deployed AI compute node?

• 31:38Molly O'Shea?

  I do.

• 31:39Speaker 00

  For distributed command and control of a AI enabled battlefield?

• 31:42Molly O'Shea?

  I feel like I'm about to go save America. That's what I feel.

• 31:45Speaker 00

  There you go. You want to play the National Anthem or anything here?

• 31:49Speaker 00

  Yeah. So to your question, it's really like, I prefer, especially in earlier phases, to buy a lot of this capital equipment. Used off the secondary markets.

• 31:59Molly O'Shea?

  Yeah.

• 32:00Speaker 00

  Cause it's like, what's the, what's the chance that our requirements change? What's the chance that we like, we think we need this size or this scale, but then it evolves and we need a different? Different version or a different model, like it's pretty high. You wanna drive off-road a little, you just wanna take, you know?

• 32:15Molly O'Shea?

  Is there an off-road track?

• 32:16Speaker 00

  There's not unfortunately.

• 32:17Molly O'Shea?

  Should we just drive?

• 32:19Speaker 00

  You can drive right up here, right up through the lobby.

• 32:22Molly O'Shea?

  I gotta go to my Palmer interview really fast, so I'll just take it into the lobby.

• 32:27Speaker 00

  Yeah, you can just bring it right around here, park wherever you want.

• 32:31Molly O'Shea?

  In the middle of the road? You don't trust me?

• 32:36Molly O'Shea?

  Okay.

• 32:37Speaker 00

  There you go. If you wanna come see the, uh, kind of one of the mini, one of the versions of Titan here.

• 32:44Molly O'Shea?

  Oh, do you?

• 32:45Speaker 00

  Put it in, up, up, flip that one down. There you go. You just did the same drive, uh, Sekwar Hegseth did a couple days ago.

• 32:51Molly O'Shea?

  Who did it better?

• 32:52Speaker 00

  I mean, I don't wanna get into political hot water here, but, um, he jumped the curb and drove it into the, into the crowd for the speech. Yeah, it was a whole thing, so.

• 33:01Molly O'Shea?

  Alright.

• 33:02Speaker 00

  Were you late for Palmer?

• 33:03Molly O'Shea?

  Well, this is what we have here.

• 33:04Speaker 00

  Yeah, yeah, this is, this is Menace. This is Menace, yeah. And then this one's Titan. These are basically the same kind of concept of this distributed kind of compute node. Bigger, more power, more compute, lets you do more interesting things for different missions deployed, but harder to move.

• 33:24Molly O'Shea?

  Yeah.

• 33:24Speaker 00

  And then, kind of in the middle, easier to move, medium blend of compute and all, and power. It's just different kind of vibes for whatever the,

• 33:33Molly O'Shea?

  Yeah.

• 33:33Speaker 00

  Uh, whatever the customer wants. In practice, what they will want is they will want

• 33:40Speaker 00

  one of these at a certain kind of hierarchy of unit, and then one of these at a lower level, one of those at a lower level, all the way down to the.

• 33:46Molly O'Shea?

  Do you like, essentially like, order them as a, as a set? Like, do you, like, do you need them all together? Do people just like them all together?

• 33:53Speaker 00

  No, I, we like them all together.

• 33:55Molly O'Shea?

  Yeah, of course.

• 33:57Speaker 00

  Works better that way, uh, to say nothing of being good for business, so we're gonna head.

• 34:00Molly O'Shea?

  How many times do people make a mistake of just buying one thing at a time?

• 34:04Speaker 00

  How it really works is like they'll put out a, um, A requirement or a bid and say like, "We're looking for a new command and control platform." And then we'll go put together a proposal and say, "Hey, we've got these different products that we've invented that we think can meet the needs." It's more of that kind of style than like they're ordering off the menu or something like that. So it's a different kind of vibe.

• 34:24Molly O'Shea?

  But Fury is your favorite?

• 34:26Speaker 00

  My personal favorite.

• 34:27Molly O'Shea?

  Why?

• 34:28Speaker 00

  I mean, it's a robot Top Gun. How's that not cool?

• 34:32Molly O'Shea?

  It is cool. They're all cool.

• 34:33Speaker 00

  It is very cool.

• 34:34Molly O'Shea?

  Hey, it's Molly. If you enjoy our interviews, check out our newsletter, sorcery.vc, where we deliver a once-a-week top deals and tech headlines email and also go deeper on our podcast interviews. Subscribe to Sorcery today. And don't forget to subscribe to the podcast on YouTube, Spotify, Apple, or wherever you listen. Link in description to sign up.