Anduril and DRA announce multi-year deal to deploy AI-native manufacturing software across Anduril's factories

the TBP and Ultra Dome. Come on. Come on down, guys. Uh, thank you so much for taking the time to come to the TBP and Ultradome. Great to see you.

Good to see you.

Two of the most athletic men in TAC. Both athletes in their own right.

Bench press world champion, bench press high school champion. What was the what was the stat again?

Uh at the age of 17.

17.

Broke a bench state record for New York.

Let's go.

That's the real news.

What What was the actual uh amount? The weight.

Nothing crazy.

Oh, okay.

What's the number? What's three plates?

It was It was 275. At the time I could do three plates.

I overshot it. I do 355. I didn't actually get to do a 335, but it is what it is.

Wait, that so that was a high school state record.

Yeah.

Yeah. I was for age and weight class. I was 17.

Amazing.

I weighed like 190 at the time.

And what's the latest on the Murf? Is this an annual thing now?

It's an annual thing. We're doing it again this year. Rolling it back. Doing

time. Uh definitely beating the time this year. Yeah. I have one 1 hour goal. So I was an hour and 12 minutes last year. So

Okay. You can shoot on 12 minutes.

Yeah. Yeah.

That's what 20% of the time.

A lot of training. Yeah. You going to come out and do it this year?

Yeah. I'll be

Last year there was a lot of talk and then uh not a lot. [laughter] So, you know, you coming out heavy,

you know, you know, when we started the show, uh, you invited me down and we did a Murf or we we didn't even do a full Murf. We just did like some training and I was completely dead. Uh, and I went down and I had like a fever for like a month. [laughter] It was brutal. I was so sick doing the show. Uh, it was absolutely

But this year, yeah, we're rolling it back. And then most importantly, we're going to do a big one in um in Ohio as a partnership with Ohio State and of the launching of our also one factory there. Uh, with everything getting up and moving, by the time the Murf comes around, we'll be just opening for production at the factory there. So, it'll be a pretty big day for us.

Great timing.

Uh, let let's rewind, introduce yourselves just so we have it on the record. I think everyone knows since you repeat guests, but kick it off.

Oh, I mean, I'm Phil. I'm the co-founder and CEO of DRA. We are building the first AI native

production orchestration platform starting with automated work instructions.

Okay.

And then I'm Matt Grim, the co-founder and COO of Andreal Industries.

Okay. And then why are you two here today together?

We are very very fortunate to chat with you guys about an excellent partnership that the folks at DRA and Andre have come to.

Cool.

Um we signed a excellent uh multi-year agreement to roll out DRA across all of Androl's manufacturing facilities.

Okay. So that includes the one in Ohio.

We'll include the one in Ohio. Yes. So I I I first met Phil about two years ago and change. Um there was a video that they sent around that went kind of viral in uh in kind of manufacturing tech nerd circles and it was a video of the an earlier iteration of their build OS uh showing how you could go from a model in CAD and that the uh the Build OS platform would then take that break it down and turn that into instructions for factory floor workers to actually put together and and assemble. So I saw the video reached out and said like hey this seems kind of cool. We've got a lot of factories. We're we're growing. We're building uh you know could we could we set up a call? So we um kind of did a couple of calls, did a couple of videos, you know, video zoom screens and then um I uh brought brought him in to do a proof of concept and we've been kind of building together for the last two years fully integrating the build OS platform into our existing suite of tools. We call the tools the the Arsenal OS that is the combination of our ERP on the back end. Um then our um our PLM that our CAD designers use every day to design the software. Our uh MEES on the floor that our production workers are interacting with and then getting uh DRO um build OS plugged directly into that which will then reduce the amount of time going from an idea on a whiteboard to a model in CAD then ultimately to those work instructions that the factory floor workers need to actually assemble the products. Got it. Mhm.

So with uh with build OS implemented, we've seen pretty wild wild improvements in the in the time that it takes to go from that CAD model to that work instruction for the factory floor. Was the number 80 something%

87.5%. There we go. It was 87.5%. It was a 12h hour redu. It was from it went from 12 hours down to 90 minutes.

That's amazing.

Pretty without without

Let's give it up for SAS.

SAS. [applause]

So how not not getting enough love.

How often are the work instructions changing? Is this is this just is this more unique to Anderal because you're developing new products, new CAD designs, and so you need to go from CAD design to work construction every day, every month, every year versus someone who's just like, "Look, I make five, five, six rounds. It's the same CAD design, same work instructions for the last decade. I'm just doing the thing that I'm doing." Why is this important for you? And and and and how does Direct plug in?

Yeah, you there's an excellent question. So, you're not you're not just a pretty talking guy. [laughter] You know, you know some things. Uh so the challenge for a company like Androl as we scale up and we're um diversifying across a number of different product lines and a number of different geographies is then you have engineers with an idea who want to make a tweak or make a change or sometimes we're responsive to quality issues in the field or something like that. They want to make a design change.

So then every time a design change gets made the current method for doing work instructions is predominantly manual. So every time a bracket or a wiring harness or a connector or a circuit board changes, you have to change the work instructions. Else on the factory floor, you get an instruction that says like, you know, uh screw bolt number ABC123 into this hole, tight into this torque, whatever. But it's a different bolt now. Y

So then the guy on the floor gets an instruction and is like, but it doesn't it doesn't fit. So then for us as we as we scale and we're specifically trying to take a faster iteration kind of faster mentality uh to to this manufacturing process reducing that time is absolutely absolutely clutch and critical. So for us it is especially paramount. Um though I will say that I think this is a problem that plagues a lot of industries and a lot of our supply chain a lot of our um other partners and vendors. So I'll let Phil speak more to the rest of the market but um it's definitely compelling and working for us. Yeah, because you've had success in like automobile manufacturing as well.

We're pretty we're pretty vertical. Last time we uh last time I came on the show, we talked about a ship building partner. That's right.

Um and makes a ton of things, planes, ships, all sorts of different things.

And what we've seen, at least with DRA, is that

our problem set that we work on, right, automated work construction, it's not just useful for just defense folks, right? Because we end up working with folks in aerospace and defense, automotive, agriculture and construction machinery, maritime, complex industrial machinery. Really, anybody that has any sort of complex assembly, you need operators on the shop floor who know exactly what to build and when.

And when it takes weeks or months to make a work instruction, a guide for your folks in the shop floor, most of the time they're operating on outdated information. And when you want to have an industrial base that is agile and robust, you can't have your operators and your technicians operating on outdated information on outdated context. And that, you know, we we often talk about the work instruction as the core atomic unit of information in the factory. Sure.

And if that is the core driving force by which your factory builds, you can't let it be manual, can't be PowerPoint, you know, can't be.

Yeah. My only reference here is if you've ever gotten, you know, like like some baby furniture thing and you got the instruction set and like if you're actually following it closely and you realize that the instructions are wrong, it's just the most enraging experience ever. And so I imagine even for like like employee morale like on the shop floor if like you're constantly dealing with just like terrible instructions, why should somebody be like

yeah, it's great if they're high agency and they're going to try to like fix it and and manage up,

but at the same time a lot of people are just going to be like, you know, just actually become less engaged because they're like, I'm not getting the right information and the resources. Why should I even why should I even try?

Correct. And and in your baby furniture example, that is uh that is not an electronic product. So you have to think of that exact problem and then layer on all the complexity of electronics and wiring harnesses and power distribution and all those extra components. So like one of the major challenges is that there's a quick iteration of chips. There'll just be a new chip that'll be faster on a different board kind of new power new power specs or whatever. So you need to make a slight tweak. The traditional way of doing that is to take all of those changes that you want to make on a very slow cadence somewhere on the area of annual sometimes uh sometimes bannual and batch all of those changes up into a single block upgrade that you then do all at once. So then for a long period of time you're deploying outdated hardware. So you want to be deploying the best that you possibly can to your customers. But to Phil's point, in order to get the best onto the factory floor and then ultimately into the hands of the customer, you drastically need to reduce the time from idea to work construction to factory floor. And [clears throat] uh and then most importantly keep all of that in sync because when you have at our scale with thousands of design engineers and then many thousands of factory floor workers, like keeping them all talking off the same sheet of music becomes a really hard orchestration problem to keep everybody talking.

Thousands of design engineers. That's a ton.

We just passed through 7,000 employees. Yeah. Congratulations.

Thank you. Thank you.

We got to bring that gong closer for you. How How are the work instructions actually evolving from a form factor standpoint? Like I imagine historically you just get like a printed out kind of I don't know big stack of papers. I can imagine

like a very thick IKEA instruction. And I think like the number of companies have come and gone trying to you know basically put work constructions into like you know like VR stuff like that. How far away are we from that? I imagine this is a question that's come up like with investor pitches.

VR is an interesting modality for displaying work instructions. Typically what we see nowadays on the shop floor before we come in uh most manufacturing engineers are using PowerPoint or Word. You know [clears throat] Microsoft that uh that old beast is still dominating the work construction market. So they just have a laptop and they're just kind of

um often times. So they will draft it into PowerPoint or Word. They'll then print it out and it'll sit in a thick white binder on the assembly station. Um if they're lucky they're updating it maybe, you know, once or twice a year flipping through say, "Oh, it's this version. No, wait. Get that binder off the rack over there uh to to make this different model." On the the the way that we think about it, right? The way that we've seen it uh on the design side of things, right? We've seen 15 20 years of an explosion of excellent software in the design side and because CAD is so advanced what we've seen is an explosion in the uh variance that people are able to design right uh think about the increasing uh complexity the different modular builds right uh you know company like Android right has modular products they have different payloads that are reconfigurable now think about that from the design perspective great we can swap out a different uh you know a bunch of different modules

now from the manufacturing perspect perspective. That means that you have to have uh when you can when you can have thousands of different uh combinations of things, you now need thousands of different work instructions where if one picture is off and one step is off, you have somebody building a completely different thing. And thinking about that from the top level build perspective, that makes sense, right? Like that's a ton of different configurations. Uh but when we're looking at all the people who make up the supplier base of the companies that make planes, cars, trains, everything in between, we're talking about companies that make pumps and valves, complex mechanical subasssemblies, right? Think about how many different variants those guys have to have. Uh now they're in deep trouble because when they lose the ability to not only have folks on their shop floor that are, you know, 30-year veterans who just know off the back of their heads, you know, okay, this is yeah, I remember how we built this last time. I don't really need instructions because all those folks are also retiring. Uh our entire industrial base and our supplier base is also at risk because we haven't codified all of that tribal knowledge. And that is a key mission of ours as well. Not just to be a rubber band around the variance, but to also codify and capture

basically like how do we used to do it? Why did we stop doing it that way? Like making sure that's maintained. So if a guy doesn't retire, you're not calling him up like you know what you used to do in this situation.

Yep. Exactly. similar to a McKenzie analyst who uh put together a uh a spreadsheet or or a PowerPoint uh that was like you know 150 slides you're like hey how did you what macros did you use what what formulas did you use for this thing um that's like another framing to think about it um except that guy retired and he's

how many products does and have now because I think most people know Fury Roadrunner dive but there's different variants with each product um how how do you think about the product footprint

to to fill to fill 's point that there are roundabouts 30 of the core products, but then there are many many hundreds of individual variants that'll come along with those. So, for example, the um the Australian military might use a slightly different radio than the US military. So, then there's just and it's not a not a hard change, but it's still a change, a slightly different bracket, a slightly different connection.

They're not standardized on a specific size. So, you need different bracket sizes and all that.

Exactly. And then to Phil's point about payloads. So for on the dive XL for example, uh maybe you want to put a very advanced sonar package or maybe you want to put a mind sweeping package or um something along those lines.

So that the actual dive is still the same, but then inside you need a slightly different bracket to hold the thing and a slightly different

plug USBC for this is this is this is a this would this would be the ideal. I think it's a long ways out though. So

it is right because standards I mean they it takes so long even just to roll out into the phones. I mean how long was it before the iPhone and the Android phone had the same charger? Clearly we need uh European style regulations to force to force us into it.

Seriously like do we like like does it need to be defined from government?

So there there there are a number of standards that exist out there specifically about um you know communication protocols and things and then this is part of why we run these international u multi-military exercises partly to kind of work through these types of issues

especially for comms. I imagine if this helicopter can't talk to that plane you're dead.

It's a it's a it's a it's a huge pain point. Yeah. So there is some standardization, but that doesn't quite get down to the chip level like in inside the the the seeker head of a Barracuda that there's a maybe a slightly different sensor that this customer or that customer wants to use. The body of the the Barracuda is the same, the wings are the same, the engine's the same, the power module's the same, all that's the same, but at the front there's a slightly different module, which then to Phil's point means then you need another set of work instructions, you need to train your workers a different way, and then that [clears throat] becomes uh very hard to manage at scale. Um, so with with Phil's uh platform, we'll be able to kind of bring that together and shorten that time and ultimately lead to products getting in the field quicker.

Sense. Uh, talk about Ohio. Uh, is it is how far along are you with the project? Uh, is uh is Trey vindicated in everything else? Sure. In all of his Ohio,

you know, I mean, uh, don't don't want to take swings at Trey here, but um, the, uh, the the Arsenal 1 project in Columbus, Ohio is is is running incredibly smoothly. We've got about the first 30 and change um Ohioans hired already. They we hired them about 6 months ago and then brought them out to California. So, they've been at headquarters where we've been building the first few tales of of the Fury line. Okay.

Um then they're coming back to Ohio this month and then we're um doing the standup and provisioning of the line. So, the ultimate Arsenal One campus will be about a dozen buildings totaling in the neighborhood of about 5 million square feet total. Uh the first building that's about um just shy of a million square feet. It's like 800,000 ft² um is is fully built. The internal is being being constructed now. The inside uh and the production line on all the equipment, all of that, test rigs and all of those things are being provisioned. And we'll start production on Fury uh this summer.

Building two is the walls are vertical. So that's that's that's going up. And then building three is immediately there after. And building four right there after. So we'll be rolling out a new building about every yearish for um about the next seven eight years. Slight slight. You said the Fury tails are being built one place. Is this is this a matter of like building different pieces then bringing them all together? Is there an assembly line? Is it like a manufacturing?

Sorry. Uh a a term term term of industry. Tail tail being the the full plane.

Tail like like a tail number on the plane.

Oh yes. Got it. Got it. Yes. Yes. Yes. So the first the first the first uh the first four or five I'm going to get in trouble for not remembering. The first four or five will be built will be assembled at HQ

and then test flown at our test range nearby. And then the next one will be the tail. Why not [laughter] whole of

the tail number the whole the whole plane but then but then from there on out they'll be manufactured.

In terms of assembly lines, are we still paying homage to Henry Ford with a with a plane that moves down and gets things added or is it a more integrated process

for for Fury's line? It will resemble something in the neighborhood of of Yes. So it move from station to station.

More like the grim method of

There you go. There you go. There you go. We'll coin that one. Um we need a gong hit for that. Perfect. Perfect. Thank you. Thank you. Um but the the line will resemble closer to that where at at station one we do the hydraulics. At station two you do the electronics at station three you do the fuel system. Something along those lines. Um and then it's ultimately the fury line will be uh 22 stations. And then where the fury line in particular gets complicated is that uh we have to do a number of um very complicated coatings and paintings on it. Uh okay. though that that's a little bit um tricky and specific.

And then of course the the propulsion element of it is just very very complicated to

talk to talk to you say testament to how far out Mr. Mac Grim thinks about has been thinking about Fury two two years ago when he initially reached out after seeing uh seeing our demo of build OS initial outreach was hey I saw what you guys are working on uh can you use it to build robotic fighter jets

uh and I said yeah I mean depending on how cool they are of course again

two years later here we are here we are

yeah and got what what a what a moment too yeah

and and val how validating this is for DRA I mean I feel like part of why you guys are coming out and talking about it together is I imagine you want your supply chain to probably get on

the rack. Is that is that a part of it? Like wanting the whole or is it just a quality part that you get at the end of the day?

Good question. So, there's a couple parts uh couple reasons why I'm here in particular to to support the announcement here. First is that um Andrew's on a mission to save the West. It's kind of the the name of the company baked into that. And it [clears throat] is my firm conviction that part of that journey will be re-industrializing the West and re-industrializing America. And part of that is getting the supply chain teams and get manufacturing organizations, production operations to be more tech forward and be more tech- enabled. And just like accounting software and sort of like traditional kind of big big software platforms, a lot of the core technology in that um ecosystem is is pretty hard to use. It's almost user hostile. It's just very difficult. So part of the mission here is that like yes, of course, we want to adopt it. Then we want the supply chain that supports us to adopt it. And then in the long run, we want more and more and more and more factories uh operating this way because if we're going to compete with um with China and the rest of the world on the large scale manufacturing, we need to get faster at this. And part of part of the the the lag in Western manufacturing is exactly that to your point about the the the work instructions of a of a kid's play set that aren't quite right.

What happens on a line then is it's like then the line is stopped at station seven because the station seven instructions. So then you just get like, okay, well then it stopped here, stopped at six, stopped at five, stopped at four because that guy is like, I don't know what to do. I can't fit this. So then extrapolate that out across industries and there's just so much wasted time and effort that gets spent on problems like this that I think can be solved with technology. It's a perfect application of where AI can be very powerful to help accelerate the development and the shipping process. So that's one point. And then the other point I want to bring up is that we've seen this very fascinating and um honestly unpredicted uh outcome of our work with DRock is that by talking to our customers so talking to the Navy, the Air Force etc about how we are going to scale manufacturing and talking to them about how we go from designs into PDRs and CDRs and MRS the whole engineering process to get something designed and fielded. We've been specifically pitching DRock solution as part of that ecosystem to help them understand exactly this question about how we get from that idea to that work instruction to the factory floor to help us accelerate. And it's been

pretty good ahead of sales

in a in a bizarre way. It's been an accelerant for our sales to help our customers see that we're leveraging this this new fangled technology this this AI because the government customers are like what is this AI stuff? How am I going to use it? And we come in and say here's how you're going to use it. We're going to use it on autonomy on your end products. You're going to use it in the manufacturing line. You're going to use it in the back office. And here's how it's going to ultimately get you better products fielded for cheaper.

Yeah.

Uh I want to talk to you guys about the labor market in manufacturing specifically. Are you guys

going had there's so much demand for probably uh type of talent that could do manufacturing or they could go work in a data center development project? Are you guys competing for for talent with the data center buildout? Is there enough people enough people excited about entering this this workforce?

So there's there's actually two pieces to talk about there. Uh interesting thing to bring up data centers. Um increasingly well rather historically the way that data centers have been built out, right? They've looked a lot more like architecture projects. And so you've seen the tooling that people use to design data centers historically um on AutoCAD, Autodesk, which is a little bit more of the architecture. You know, they're doing it in BIM which is like

building a building.

Yes. Exactly.

Throw a bunch of racks in there.

Historically, yes. Uh but now something that we are seeing as a company uh through the companies that we're working with um increasingly more data center companies coming to us and saying hey we need to start building at scale the same way we produce electronic systems at scale. And so I you know we are seeing a massive migration of data center producers and data center builders away from traditional architecture oriented uh design and build systems towards the more manufacturing oriented design and build systems. And with that they're saying hey uh we are going to have to build data center systems all over the world and we cannot have folks not having the context of what to build. Can Durax work construction software be useful for this? And so of course we say yes. Um and so that's one thing about the data centers uh just as an interesting anecdote that might be interesting for folks to to know about um that we are also trying to build data centers the same way we build cars and planes and everything else. Um and then on the labor question um another thing that we think about is radically reducing the barrier to entry for net new folks into manufacturing because you know picture is worth a thousand words but an animation is worth a million right earlier you were asking about how do we change the modality of the way a person interacts with a work construction

with our software it is not just a picture not just a photo not just a printed out thing it is a 3D interactive animated uh dynamic work instruction that a oper operator technician can interact with, right? Pictures worth a thousand words, but animation is worth a million. And so, because you were able to have less skilled folks potentially take the roles of operators and because you're also codifying and aggregating all of that tribal knowledge and context.

Yeah.

We're seeing that the companies that we work with are actually able to hire less trained technicians. So, it's easing that burden and pain for them.

Yeah. On the on the Androl side, uh I' I'd add a couple points there. First of that is is that on the data center side in the construction market it's becoming pretty contentious just because you know big huge builds need big huge bulldozers and concrete layers and um electricians and etc. So on the construction side

you guys are basically competing for the same kind of firms for the time of firms that are getting offers from data centers that maybe have a different

so on the construction and build kind of side of things. Yes, it's that's certainly getting competitive on the actual technicians on the factory floor. we haven't seen pressures from data centers because the the the dot I would add is that these data centers while very big and very powerful tend to employ very few people. There's not that many people engaged in the actual day-to-day running of the place.

So for for us like on the technician side we don't see much competition from them. And where we are spending a lot of our effort is investing into the local communities including in Ohio through partnerships with the local universities and trade schools and community colleges like that kind of ecosystem to work with them on the types of skills that we want to see new uh new hires have whether that's additive manufacturing or welding or particular CNC programming or just like electronics assembly whatever that skill set is helping design curriculums that can then prod produce graduates of those programs that are ready to come get a job at Anderoll. So

for us it's kind of a kind of a long play cuz you know we invest that time up now and those graduates won't come out for a couple years but um in the long run we think we'll will lead to a pretty fertile supply to say nothing of being good for the communities that we're engaged in.

What are you seeing within the armed forces on the self-repair side? Uh there's been a couple press releases about being able to repair drone or something on the fly. I remember I toured a toured a a aircraft carrier and they had a machine shop and there was a a sailor in the machine shop who was talking about if if something breaks on the ship they build they they will go and mill the bolt right there. They don't go and buy it because and he was very proud to say, you know, we're saving the taxpayer. They want to charge us thousand dollars for this bolt. I made it for five bucks. Uh that seems like a really positive thing. Does that play into automated work instructions at some point? some digital solution and then just what are you seeing on the overall repairability from the systems that you'll be deploying and are deploying. So, uh, yeah, uh, funny enough, maintenance and repair is a very, very common next question when folks ask us about work instructions, right? Most of the time when they're saying, "Hey, uh, awesome. You can now automatically generate my work instructions for me. Uh, I can now have operators and technicians interact with it in a 3D manner." Awesome. Now, can you do maintenance and repair instructions? And our answer is, of course, yes, right? MRO is a natural uh, inverse generally of actually the initial build of the thing. Uh, and so part of what our roadmap is, part of the things that we're working on with a variety of different folks is is automating maintenance and repair instructions. So, we're very big proponents of, you were mentioning, uh, the right to repair, not just of, uh, you know, tractors, but of drones out in the field. Um, it's a very big thing that we've been talking to a lot of different folks about, and it's something that we're a big proponent of because we don't think that these should be two different uh, domains of instruction sets. Um, context should live on the model. context should live in the hands of folks who are actually designing and building those systems and repairing them. And so you should at the time of making your first instruction set have the ability to codify all that context into one core instruction set. Um and maintenance and repair instruction should fall naturally out of that. And that's it's a lot of how we think about things.

Yeah, it makes a lot of sense.

Yeah. from from a policy perspective where I think the lines get a little bit blurred is that we're talking about a bolt or a bracket or a fixture or something that you can carve out of metal makes makes obvious sense in one area

and then the next area would be some somewhere in the middle of you know rewiring or reconnecting you know an additional battery pack or something like that that's kind of in the middle and then all the way on the other end is like all the way down to the chip level

chip broke you're not going to fab it

yeah you're not you're not laying laying silicon silicon chips out on an aircraft carrier or something right so there there is a spectrum there at which I mean, at some point on that spectrum, you say like, "Actually, I do need to call the experts to help me on this,"

right? But, but, uh, in my opinion, the the the default of saying, "Don't touch it. We'll come repair it. Ship it all the way back to our factory." That's just not going to fly in any time of of of conflict. Um, cuz like we do need our sailors and our airmen and our soldiers in the field to be able to at minimum replace the consumables, you know, the brake pads, the tires, etc., but then also to do kind of the next level of repair into, you know, replace a fuel pump or something along those lines. When it gets down to the chip level though, you're like, I'll probably send you a replacement module and then maybe the work instructions say take bad module output, new module in, reconnect these cables. Um, but the in general, yeah, we want to be as as friendly as we can be to this forward repair concept because in a time of conflict, you just think like I need the thing now. I need need to go. you said brake pads and my uh mine went to motorsports because I know Andrew Rolls is supporting NASCAR and then I was curious are you guys working has there been any inbound from any racing organizations that that are make having to make parts on the fly and are in heavy R&D and maybe they only make the part like maybe they're making different variations of the part but they're not actually doing it at scale because it's not a manufacturing business.

Any any opportunities there? So, funny enough, uh I cannot confirm or deny explicit names. Okay. Uh but we are actually in conversation with several motorsports teams uh who are actively evaluating us for use cases on and off the track.

Yeah. Probably a like a very different workflow, but uh so many learnings and just great for the brand and whatnot work there.

Exactly.

Yeah. Uh how are you reflecting on the DoD or Department of War budget potentially being increased?

I still say DoD too. I'm trying. I'm trying. But it's, you know, years

um I mean the the the the post saying that we might go to 1.5 trillion uh that seems like that was not Ander's original mission. Andrew was all about uh doing more with less. Um but at the same time if it accelerates the mission that could be good. How have you been processing it internally?

Well I mean I think any anytime a customer says they want to spend more on the things you make is generally generally a good good thing for us. Um that said exactly to your point like we would love to see these um this increase in funding if it happens of course it's got to go through Congress and all the usual but if it happens um be spent smartly and on um new technologies new approaches and not just kind of funding the old way of doing things in the old kind of um legacy model that in our opinion hasn't worked for decades. So, um, yeah, of course, you know, customers spending more money is good for us, but, um, we're going to be lobbying pretty heavily to have them spend it on more AI powered, you know, weapons of futures, autonomy, um, that kind of construct. Um, and especially on a on a different kind of contracting model, more towards the cost plus type of model. Sorry, away from the cost plus type of model to the firm fixed type of model that that we like to do business with. So, um, well, still a lot to a lot to digest there, but we're in general pretty excited as part of

that initial like flipped model of like, you know, like raise venture, take risk capital, build a prototype, build something, then show up to the Department of War with a concept or something that's maybe ready to be fielded and then say buy it. Uh, has that held as long as you thought it would? Do you think it holds forever? Because at some point, if a customer shows up and says, "I'll pay you to do the R&D, what are you going to do? Turn them away?" But at the same time there's some you know there's the financial hazard that the firm was founded around. So how's your thinking evolved around the whole flipped model and doing R&D on your balance sheet?

Yeah, a couple of thoughts there. First is that um we spend an inordinate amount of money on our own um internal research and defense IRA type uh research and development rather

um spending just an absolute inordinate amount of money and we're grateful to have the support of our investors who believe in this business model. So for us, we haven't not at all seen a like let's just shift back to that way. It's a lot capital, you know, less less risky, capital safer, all of that. Um, so no, I'm not surprised like this is to your point, this is exactly the kind of thesis, the mission of what we're trying to do. If anything, we've been very happy to see venture capital funding to the se uh to the sector increase to fund more companies approaching it this way because I think that for us to win a great power conflict, we're not going to need one and we're going to need five or eight or 10 of them. So, we're excited to see, you know, new companies get started with this kind of mindset and um and go out and chip off parts of the parts of the uh ecosystem here.

Yeah.

Um the last point I would say is that uh to date something in the neighborhood of um 85 90ish% of our uh revenue has not been from these kind of cost plus traditional like I like customerf funded models. Like the vast majority of it comes from us selling our product to our customers at a fixed price. Um and then um with a recurring bit of recurring revenue tied to that for kind of think of it like an extended warranty kind of concept where yeah we're doing the maintenance, we're providing the support, we're doing the security upgrades, the patching, the all of that which from our perspective really aligns incentives for us to deliver the best possible product because to be blunt like we will have a higher margin if we don't have to send repair guys out and keep replacing things. but then also gives the customer a more reliable product and then importantly gives the customer predictable pricing. So they know what their total four year fiveyear cost of ownership for a product will be because they know that within that number we're responsible for anything that breaks any sort of problems or maintenance that needs to happen

that aligns everybody and we're seeing a broader acceptance of that business model within both um the DO and with DHS.

So we're pretty pretty excited to see those messages landing.

Yeah.

Give us an update on the secondary wars. [laughter] Oo, spicy.

Are you the head of investment relations or

I'm not. I'm not. I'm not. Um,

Alison Lazarus is our

the guy she calls to lay the smackdown.

To be honest, at some point there's going to be some enterprising young SPV slinger that's like, "If I make a really heinous memo, Matt Grim is the Grim Reaper is going to come to Wait." So, so, so, so give us some backstory first and then

Okay. So, first I would say that uh our investor relations is one by run by this incredibly talented woman named Allison Lazarus. She's great and she works closely with Trey on kind of managing the cap table and all the inbound and interested investors and all of that. So, um not really my not really my turf anymore. That said, um I've had a longunning battle against what I would call the wild cats in the secondary market who have absolutely zero respect for any sort of uh control of the cap table like who our investors are, information rights, any of that. Um and they're just out kind of slinging offers at early investors, early employees at share prices that are completely insane. And those people don't care whatsoever about the impact that has on the company, whether that's around 49A valuation or RSU pricing or any of those kind of internal dynamics. They don't give a [ __ ] So, they just kind of create headache for us by then us having to field this inbound. So beyond the headache, the point that I'm really worried about and I think the SpaceX IPO will be very very interesting because this is going to be a situation where there are, you know, the tide goes out and there's going to be a lot of people without trunks on that's just they uh people have been out there slinging in our case uh access to and roll or I have limited access to their next round

at some insane price that is nowhere close to what the actual price is as set by the market. Sure.

Uh and then people are buying

and and that's not even factoring in the fees which are sometimes crazy layered fees. So there was there was this one situation of um someone slinging an offer around that was an SPV into another SPV that was buying a chunk of an investor's an early investor's holdings.

So there's like layers and layers of fees and carries on top of that at a price per share that was something like 70% above what the last round traded at. So, it's just an insane thing where they're pitching access to a deal they they don't have.

And I know for a fact that there are there was a recent indictment announced in New York around a case where someone was uh was outslinging and SPVS.

No way.

That did not have access to and the guy was just pocketing the cash and then he was just indicted and arrested at JFK. Um, so like this is a thing that happens. And the reason I bring up SpaceX is um, not to create more work for them, but that the the world of SpaceX IPO because the legal fees of unwinding all these wild, right? Because there's there's just there's a lot of this nesting and SpaceX is a fantastic company that's been in business for a long time. So there's so many of these nests of who owns what

these types of like main street investors. It's totally possible that they're thinking, well, I got into SpaceX at 200 uh when you didn't and then it goes out at 1.5 and that's my retirement and then it goes out and you're like, where are my shares and the guy skipped town just out whatever you put in and not only did you lose the whatever you put in, but you also lost the mental of where you thought you were and what you were planning. You might have been planning a certain lifestyle

course and the opportunity cost of deploying that capital somewhere else, right? It's a it's a it's a very big problem. And um I've used this joke before, but uh my my joke is that how many investors in America think they own a chunk of SpaceX when they're actually funding their ex roommates's boyfriend's coke cabb in Miami?

And it's [laughter] like

a fair amount probably a nonzero amount, right? So what do you think you think the solution is? Can the can the uh community police itself? It seems like obviously like you have a legal team that's working on this stuff, but there is an element of like going direct like you need to get the word out that this is happening and that's important too. But

so this is so this is why uh I've we as a company have been more public about this is because we're specifically trying to send a message to the market that's like

we see you, we're aware of you

and like secondaries of course have their purpose. Of course there are early investors who need to liquidate funds or fund life cycle problems. Of course

so this isn't a blanket comment that all secondaries are bad. They're not. They're not. There's a category of these people who just have like really no respect and are basically just fraudsters, huers. And what I think needs to happen is two things. One, the investor community needs to be better at policing ourselves around just who these bad actors are. Um, and kind of shining a light on them. And then second, sadly, I think it's going to take law enforcement action, whether the SEC or the FBI, you know, thacking enough of these folks that that community starts to see like, uh, this is this is this isn't good for me. And it just kind of puts a damper on on on that side as well. Mhm. Mhm.

Last question for me. Uh, Adam Porter Price has been on a tear doing deals, buying companies. Y uh, as COO, the As COO, what is the key to a great postmerger integration?

The That's a great question. Um, and and an often overlooked uh, part of doing these deals. The head the deal makes a headline.

Adam busy popping champagne and then you got to get to work. Adam logs more more airline miles and more and more iPhone minutes than me, than Brian, than Trey, than anybody. He's because he's constantly out there, the best of the biz doing this. Um, so from a post merger perspective, um, there's a tickytacky piece of it that is important, uh, which is payroll, benefits, equity ownership, um, you know, all the 401k transitions, the email has to migrate, like

but making it sure it happens on day one. There's no, oh yeah, you'll get your paycheck, but it's a delayed a week. That's annoying. No one wants that.

Correct. Nothing will burn your trust with a company faster than screwing up their first paycheck. So, for us, it's a thing that we try to lean as heavy as we can into that employee education and systems transition up front.

Um, it's not possible to transition like ERPs and years of design software. That takes usually, you know, 6 to 9 to 12 months on the long end to get those systems fully working. But like the guts have to work first.

Yeah.

And then second is I would say is a uh cultural alignment. It's just like how we pitch our products, how we price our products, how we plug into the kind of go to market

ecosystem and engine like we have to we have to get that right.

Uh and then then the third piece I would say is trying to get some wins pretty quick.

Uh because if you're if you're an acquired company and you get integrated and you're like, "Oh, who are all these people? Is this different? I liked it when we were smaller." Yeah. Oh,

maybe they're in scrappy mode and then they're like, "Hey, I'm in.

I saw the campaign don't work at Andrew. Why would I [laughter] just don't work?" while I'm so kind of aligning with the what are we what are we trying to build? Why are we building it? What's the goal? What are what are the deliverables we're trying to hit? Like getting that kind of alignment for folks to see kind of see the destination, see the lighthouse on the hill and say that's where we're going and um aligning folks to that. And it's it's those second two that are that are really hard because that's much more kind of personality based and kind of relationship based. So it takes a lot of investment from the leadership team, takes a lot of investment on um the engineering team especially, you know, how are you designing your products now? What's your release cycle like? How do we get that into Andrew's kind of way of doing business? There's just a a big transition there. But

yeah, you got to migrate.

It seems that's exactly right.

There's been this uh maybe kind of sentiment or joke uh shoot for nice even if you miss you'll land at Anderoll. you know, this idea of like, you know, maybe maybe I don't build a multi-billion dollar company in as a new like neoprime, but I have, you know, maybe like the the the not the next outcome as I land at Anderol. Can you reality check that because I feel like in practice?

Yeah. Yeah. Yeah. No, no, no. And and and I'm just saying like I think it's like for founders to understand like what is the actual bar because you guys every every company that's built a great product but maybe the business side isn't like fully working yet and they're thinking like okay like my product's great, our team's great. I bet we could fit in there. Reality check that.

Yeah. So, uh the reason I'm pretty dismissive of it is the um that implies that Andrew will acquire anything or any company that's struggling or not

because Adam's logging all these miles but he's not exactly doing a deal. No, we do like two a year on average, if that, right? So, he's he's he's saying no a lot a lot. Um

but uh what we look for in acquisitions is really a um a sweet spot of of two things. One is a very strong engineering team like they have to have something that's very compelling and very unique and very interesting. Uh and then the second is a area where where we can add a lot of value or accelerant or fuel to their go to market.

How do we how do we speed up the timeline by

Exactly. And you guys talk a lot with defense founders and you've had a lot a lot on and folks will consistently talk about how hard it is for the go to market engine and there's a lot of nuts and bolts around that that get underestimated around what formal government proposals look like, how to do formal government contracting, security clearances, the government relations, kind of lobbying side of this to be to be a player on these bigger deals. Then um relationships within your customer that are so you're not just pitching you know the guy flying the drone out in the field. You're pitching the whole chain and how it fits into their ecosystem, how it fits in the program. There's a whole lot of that infrastructure that to be blunt like we're pretty good at. Like we've gotten pretty good at this. So when we find a company that has an incredible team building a really cool product that's like really unique and interesting but is maybe struggling there. They haven't seen kind of the ramp that they want and we see that and say hey look we can help with capital obviously

we can also help with this go to market engine we've built this legal infrastructure this facilities infrastructure this government relations infrastructure all of that will then take your product and can help really accelerate and get it deployed both to the US and to friendly governments around the world like that ends up being the sweet spot. So the the reason I would say that uh that that the base question that you started from is just completely off base is like we're not going to go look and try to find a company that like isn't clicking.

They don't have a great product. They don't have a strong team and they haven't sold a lot and then be like yeah let's go get them a couple deals a year. It's like usually like you're going in and being like we want to buy this company less like hey like things aren't working that well we should try to sell.

Yes. Exactly. Exactly. Exactly right. And and and the last point I would make on that is that uh to date uh we've done exactly zero turnarounds. And I think this is a kind of a a mental model for us is that like in general we don't really want to be in the turnaround business of take a take something that's struggling and come in and fix it and like deploy our team of operators to go fix it. Like there are plenty of PE firms who are very good at doing that. There are plenty of other organizations who are honestly better positioned to do that.

For us, we really look for that strong team, strong product, and then where we can really accelerate sales and and and go make an impact.

Very cool. Do you uh do you have a favorite? I know. I don't want to put you in hot water.

There's no chance I'll answer that question. No,

don't put I I love all of my children equally. Yes. Um excellent.

So, which one did you log the most? [laughter]

I think we know

there's a there's a there's a great company we acquired um in in Ireland um

that was called uh called Class. Um and they make these absolutely incredible uh computers and networking.

When I saw that, I was like, I want one of those.

Yeah, they're they're they're great great devices. great team. Um, but it's but it's headquartered in Dublin, which is very far. So, uh, a lot of lot of time spent over there. But, um, and that acquisition's gone gone absolutely swimmingly. Couldn't be happier.

Well, thank you so much for coming on the show.

Thank you guys for having us. Appreciate you.

Hang out for a second.

Leave us five stars on Apple Podcast and Spotify. We will be live tomorrow at 11 a.m. Pacific. Sign up for our newsletter at tbpn.com. And thank you. We will see you tomorrow.

See you tomorrow, folks. Have a great MLK day. Thanks for hanging out.

Goodbye.

Cheers.