#38 - Olaf Hichwa, Co-Founder & CTO of Neros

26:42Building products that your customers like in every other industry is, of course, like it's... Jeff Hazers is customer-sessed as an example. In defense, somehow that's not standard. And we're really, really end user-focused. We want to build a company for the users. Yeah. I mean, I imagine even from kind of that perspective, the end goal is to equip militaries with superior capability. And so if you're focused on just some guy that's signing paperwork, that's not exactly – he doesn't know necessarily what the frontline people need. For sure. For sure. There's a disconnect. And it's not to say all people in procurement are incompetent or stupid.

27:27You know, most of the people in procurement were end users themselves. but no matter how skilled you are being in procurement it's very different than if you're on the front line if you're training if you're that kind of thing and uh you know one of the things they do very well in ukraine is they just give soldiers money uh there's like you know these like amazon style storefronts where um you know users can buy drones um and they don't get a ton of money right there's a lot of people in the military but giving end users that kind of flexibility i think is really really smart do they basically get to choose what equipment they're buying?

28:00A little bit. So there's a lot of reasons. As an example, providing a counter example to my point, if I gave end user a million dollars, they would buy a Chinese drone because the Chinese drones are better. So there needs to be some, and if we just completely rely on China, we will lose for sure. And then if we give every end user a million dollars, they will all buy different drones and none of them will work together. So there's limits to this. But one of the things Ukraine does really well is they have good lists. And Mihalofetorov, who's someone in the startup space and someone I look up to a lot, I think someone who's really on it has been a big pioneer of this in Ukraine.

28:37But they have these web-based sites where they list the various drones that have been approved because they're safe enough to carry an explosive and they're not made in Russia or something like that. And users can buy those drones. And then also, if they get kills with those drones, if the drones work, if the product does its job, they report that back and they get cash incentives to buy more. And so you get this like feedback loop. It's just like Amazon where there are reviews. And I think there are people pushing for this in the DoD. And, you know, I see the value of more traditional procurement as well.

29:12And it's not to say that we need to give up on traditional procurement, but I do think these kind of like Amazon style storefronts are super, super cool for defense and definitely part of what is going to help us rebuild our industrial base and get to parity with China for drones. Yeah. I also kind of imagine like because you're trying to focus on a fully American supply chain, you could potentially probably like hamstring yourself short term on scaling just because you want to build it in America. So how does the tradeoff in your mind work? Like you want to maximize number of drones built and like iteration loops and everything while also like building this entire thing that hasn't existed previously yeah we could build so many more drones than we did do today if we you know production line here it's not that big it's six desks long on each side um and right now not even all the desks are full um and then you know look at all the people that exist outside they are in general like designing the components and doing the you know um stuff to feed the line.

30:18If we made the decision early, early on to do what they're successfully doing in Ukraine, just buy parts from China and focus on assembly, but hire a lot of people that do these manual solder joints, that use these Chinese chips, that use these Chinese components, we could produce way more drones. But that's not what building the American drone industrial base is about. Building the American drone industrial base is about obsessively removing every adversarial component in your system as possible. And that's about the humility to realize that no one has a fully adversary free supply chain. It's literally impossible given the way world politics, international trade works.

31:06But having the humility to realize you are dependent on kind We have Chinese chips in this system, and we're fighting like hell to remove them, and working extremely, extremely hard to move as much of the supply chain step-by-step to allied producers, ideally American producers, as possible. And also being transparent with our DoD partners about it. A lot of people come out and they say, when they get sanctioned by China, they're like, oh, joke's on them, we don't have any Chinese parts. And it's like, you know. You do. You should double check that. And realizing that the customer is your friend in this instance, right?

31:47America, the DOD wants to help us get rid of Chinese components. And if we're transparent about like, listen, we have probably the most China-free drone in America, and we've worked so hard to remove each part. We're making tremendous progress on de-China-fying the entire stack, just work with us to procure this. We've had great success, creating partnerships and just being transparent with the government in that sense. I also imagine, wouldn't this not only be the most China-free drone in America, but actually probably the most China-free drone in the world? Probably. Yeah. I don't get to see exactly what happens in Iran.

32:27That'd be a place I'd look for a China-free drone. I don't get to see exactly what happens in Russia that much. But yeah, I do believe this is the most China-free drone in America. I'm not aware of another drone on the market right now, as an example, in the US DoD market, that has a strike capability, precision strike warhead with a custom-designed flight computer, custom-designed BLDC motor driver, custom-designed command and control anti-jam radio link, and a custom-designed video link. That combination of vertical integration is extremely hard. Yeah, it's just not present. And I'd love to be proven wrong there, but I've looked pretty hard and I have not seen anything available right now.

33:13And that vertical integration is how you do Chinafy. If you're buying components off the shelf, you're just trusting that they're China-free and it's pretty much every component I've taken apart has ended up being somehow linked to China. Yeah. So I know that for like, you guys start out with 30 drones. You go to Kiev. of you come you get a bunch of learnings you come back um at what point did you you know i think probably focus is the most important thing on just like figuring out how to make one thing really really well um how did you decide to like start building the longbow and crossbow we built the ground systems and really the full the full kit based on our success criteria being successful missions.

34:00And do you want to explain what they are? Yeah, yeah. So Longbow is our electronic warfare EW resistant ground control system built for maximal range. We also have another variant that's Crossbow right there. And that's a more lightweight. You lose a little bit of anti-jam capability on the command and control link, but we still push that out. We can consistently push that out to 25 kilometers. So still some pretty impressive radio performance in there. So these These two base stations are the emitters. They're what put out the RF energy to control the drone. And then you connect a long cable to a more secure position, a position that's not being given away by RF emissions, usually underground, to actually operate the drone via a wired link, a hard RF-free wired link to here.

34:48And we realized pretty quickly, seeing our drones go to Ukraine, that if we just made the drone, they're going to run into an issue, or they're probably going to have to use a Chinese handset, Chinese controller, Chinese radio. But even then, they're going to configure it in a way that's not optimal for our system, or they're going to run into supply chain risks when they have to procure the ground station, or they're going to have a production issue, or the shipment's going to get delayed so that the drones get to the front line before the ground station does. So we found that missions were not being successful because we didn't enter that market.

35:21and that's when we were like well we have to make the full system and you know i think that can extend extend much further um you know we need to make the the backpack that carries the the drone it sounds small but it's actually super super important we need to think you know fully what was the process for that well it's just understanding that you know when we go take the system out uh we have to carry a cart and like you know five guys need to carry the system out originally that's not going to work right or you know when when uh soren and i go to ukraine like We can't reasonably carry a full system to Ukraine.

35:57That's probably a problem. I don't think the end users can reasonably either in that case. It's just thinking. Again, thinking about their experience actually using the product. Exactly. Putting yourselves in the shoes of the end users and thinking, what do I need to do to achieve the mission? So right now you're producing like 2 ,000 a month. What are the next steps? for getting to like$5 ,000 to$10 ,000 plus a month? It's a supply chain hurdle. So we're right now buying to$4 ,000 a month. And that inventory is back there. That cage is our very, very precious stockpile of drone parts. There's a lot of parts in there that may be some of the biggest stockpiles of drone parts in America.

36:44And right now, we are ordering to$4 ,000 a month. There's a big lag in how you start ordering the parts. That means at this scale, the order goes to the chip companies, which get allocation in a semiconductor fabricator probably in Taiwan, which then go to a packaging facility. Those are usually in China, but our supply chain is based in Malaysia and elsewhere. And then the packaging facility sends it to your PCBA contract manufacturer, and then they load the systems up. They wait for the boards to blah, blah, blah, blah. But there's a big delay between like, you know, placing the order and seeing drones come off the line.

37:20And that's really tough. What's the like time delay? Is that a few weeks? Minimum six months. Six months. Yeah. Wow. Yeah. For these, but usually longer. For large, large volume orders where you are ordering at the semiconductor level, it's long. So do you have to like plan out a few months and like maybe even a year or more on guessing how much you need to be producing a year out? It's a pretty crazy game of death. Yeah. We sell our system for around$2 ,000 at low volume. And at those kinds of numbers, it's a lot of money. Yeah. Well, the other thing, too, is it's hard. I mean, if you have a six-month-plus lead time, it's really difficult, I imagine, to iterate.

38:05It's a balance between staying flexible, which is something Ukrainians do very well, and giving the supply chain team hell based on the fact that the design is freaking changing. How can they order a year in advance when the design is changing? And flexibility is something we pride ourselves on. And we do a lot of neat tricks. Like as of recently, we separated our part number for the PCB from the part number for the bomb, the chips that go on the PCB, from the full assembly, which includes a test procedure and firmware and all that. That way we can order the boards way in advance, which probably won't change, but the bomb itself, the bill of materials, the parts, that changes.

38:42We change our RF matching networks, as an example, and we could potentially go to new frequencies based on that. So yes, life is way, way, way, way easier when you have a locked bomb, but we try and stay flexible. And I think that's the secret to mass manufacturing drones. Of course, a lot of it is in efficient assembly and the incredible technicians that put these systems together, our work instructions are designed for assembly, all that, but that's kind of a given. And then the tricky stuff is the supply chain and the demand forecasting and putting pretty freaking big bets, pushing the chips into the table, especially when DoD demand is not as high as it could be for FTVs.

39:23We still sell more than half the production capacity is not going to the DoD. And we closed a 6 ,000-unit production contract with the Ukrainians or for the Ukrainians through the European coalition of countries before we even got 1 ,000 units ordered in America, which is... As far as I know... I mean, if you just think about it, it's pretty backwards that the Europeans are paying for American production capacity when the Americans won't even pay for it. So we've had to stick our head way out the window here to invest in this capacity, but we believe it's the right thing to do. We believe that building production capacity in America is what we have to do, government demand or not.

40:05and we're going to do it. Yeah. Has it been kind of difficult to like educate the American kind of government leadership on the necessity of actually having like a stockpile of these FPV drones and like actually building it, make sure the supply chain works here without a bunch of Chinese made parts? Like how has that education process been? It's been tough. American, like the DOD people we've talked to understand the importance of production. They understand the importance of supply chain. With that being said, there's a lot of forces working against us. uh and like early on you know all of our requests for uh selling fpvs just fell on dead ears no one wanted to buy them and you know how long was that in 2023 yeah um and really part of 2024 and now you know it was like a year yeah hey we struggled to sell any fpvs to america for first year of our country company um uh and that was what do you think made their like switch they just sought the evidence 70 of all casualties like that that is really never happened before as far as i know i mean that's that's you know up there with the machine gun uh in in in world war one and you know we're starting to see it uh we we've um you know having having this kind of volume has has started to get a lot of u.s traction and you know right now like i said we sell a thousand drones a month that that end up in ukraine through europe and the other 1 ,000 in our current production capacity is going to America, which is great.

41:35I'm very, very interested in supporting the American drone industrial base. I think it's taken quite a while. It would make a lot more sense if America led the production initiative, but we're starting to see change. And based on the data coming out of Ukraine, based on the end users liking the system, based on our live fires being very successful, real explosives on these systems being used by end users, based on a lot of factors, our demand in the DoD is increasing. And there's a lot of believers in the government that understand that you have to buy these systems in mass, in huge volumes, at low prices for this to work.

42:14You cannot buy quarter million dollar loitering munitions and expect mass manufacturing to really, really happen. So we're starting to see the shift. It is definitely slower than it could be. Ideally hopeful. Yeah. So I know for the actual drone itself, it's like fully built with Neuros. And then I know that you, for the, like the payload, you're using like Kraken Kinetics. Eventually, will Neuros build the entire thing, including the payload? We're, we're. Or is this something that you want to even talk about? No, no, no. I have to talk about it. It's really important. And like Brian Davis from Kraken, Ryan Hayes, they're two incredible, brilliant individuals.

42:52And I spent a lot of time at their facility. And I think that, you know, I talked about how there should be 50 drone companies building a million drones per year. I think that when you have a monopoly in the industrial base, bad things happen. And I think that part of the reason these like legacy defense systems are so ineffective is their lack of like modularity and willingness to integrate with other payloads. So I think it makes a lot of sense to have a modular payload system. And, you know, the work that Kraken has done to make a, you know, firing device that is safe, a bomb that works, and the training to have the end users use them effectively, the integrations between our platform and Kraken makes it an incredible, incredible system.

43:44And I think that's what we have to optimize around. And both Kraken and Neuros will do whatever it takes to build an incredible system. But as far as I can tell forever, I think it's going to make a lot of sense to have a very tight partnership between our two companies. And I'm not an expert in bombs. I'm an expert in FPV drones, board design, electrical, that kind of thing. I'm not an expert in bombs. And I think we can share a lot of information, but having that symbiotic relationship and building production capacity in two companies across the American industrial base is a really, really, I think it's a good strategy.

44:24Yeah. I know that for this, you can kind of get to some massive volume just based on the FPV drone product. But do you think that there's going to be any future products that seem pretty obvious to build once you already have everything scaled up and working? So what goes into an FPV is a sensor suite, usually a camera. It could be a thermal imager, something like that. A flight computer, so something to stabilize the system. A motor driver, so you're going to turn motors. And a radio suite. And if you think about what goes into a fixed-wing, luring, ammunition, it's the exact same thing, right?

45:00It's sensor suite, flight controller, motor driver, radios. What goes into an unmanned ground vehicle? Again, you've got to turn the wheels. Motor driver, radio, you've got to talk to it. It's the same core set of components that makes virtually all unmanned robotics. And I think that FTD is a cool place to start because you can build a custom radio suite like we have in America that's anti-jam. You can build a custom fly computer in America. And you can do that all in-house. You probably could not build a core fly computer for a fighter jet in-house on a million-dollar free seed. But it's possible to do it with an FPV.

45:40So the FPV market is a great place to enter, to own that component suite, and then readapt them and hone that component suite. Because there's nothing that beats real-world testing. We knew freaking nothing when we did the first revision of all these electronics. And every part of it changed based on Battlefield's feedback. So being able to hone that core set of components and then kind of build the frames, the air frames, the externalities around those core components to achieve different missions, you know, air vehicles, ground vehicles, you name it, semi-submersibles maybe, I think is in our cards long term.

46:13Really? You think you're going to be hitting like submarines? I mean, that's way out. Way out. Way out. But I believe that the, you know, hey, submarines have basically flight computers, you know, little real-time processors that tell it where to go. They have communication links, which are very, very difficult and, you know, maybe fiber optic potentially. with. They have motor drivers, that's for sure. And that's like the furthest from an FPV, but there's still a lot of overlap. And then much closer is, hey, let's just take these four motors and rotate them and put a wing. And then you've got a longer range fixed wing system that I think is in our roadmap.

46:47With that being said, as you mentioned, focus is king, right? Across the entire business, you have to maintain focus. And we've got a big problem of producing a million drones. And, you know, I really want to stay focused on on winning that before we spread ourselves thin. I want to know, like, what is the biggest piece of feedback? You know, I imagine that even though it kind of sucks that Ukraine is at war, it's very positive from a like feedback angle where you're actually getting real world data on the product, right, where you wouldn't if it was just like being playtested. What has been the biggest piece of feedback that you've gotten or information that you wouldn't have expected?

47:25Yeah. It's a good question. I remember one of the first revisions that we made in somewhat volume, Charlie, of our system. It was the first time we had an injection molded part, as an example. We received a memo. In Ukraine, they have a pretty solid testing regimen where it goes to these test facilities, and they put together an assessment of the system, and they do a ton of tests. Apparently, the government has spectrum analyzers, and they put the vector network analyzers, and they put the antennas on, and all sorts of impressive technical tests. And we got back that Charlie was not fit for battlefield use.

47:58They were like, this doesn't meet the criteria. You need more powerful radios. Like the radio is not going to survive jamming. You need, you know, they wanted a better camera. You need more range. It was too heavy. Problem, problem, problem. And definitely like seeing that document, that hurt, right? And, you know, we brought Charlie to Ukraine. People were using Charlie in Ukraine. Like, you know, it was a functioning drone, but seeing the like harsh feedback and they don't hold back, right? They're like, this is stupid. Seeing that harsh feedback, that not fit for battlefield use, that was impactful.

48:32And we sent that to the Slack channel and it was a 10 person company back then, but definitely rallied the troops a lot on like, hey, let's make this better. And we've received a bunch of those, a bunch of these documents and they get evaluated all the time. And recently we received, a few weeks ago, I got another one of these documents and it's fit for battlefield use. They still had some feedback for us. And some of it's like issues with the platform. Some of it's like, oh, well, we just didn't communicate that feature to you. Like if you do this thing differently, if you employ this tactic, then you don't have to do that.

49:11But it's been super vindicating to see that difference between not fit for battlefield use and then being fit for battlefield use. And then maybe even more, getting to hang out with the end users You know, this past Christmas, instead of going home, I went to Ukraine again to see some of the users that got our platform and had a very impactful dinner in Zaporizhia with some of the folks using our system and just talked through all the component choices, what they liked about our system, what they didn't, and then got to see some of their operation out there where they prepare the drones, where they build the ESADs and all those kinds of parts and understanding like, hey, this is where your platform does well and this is where your platform is shit.

50:01And like accepting that this is not the best drone in the world. The Chinese have 100 ,000 people working on FPV. America has 100. I kind of love that they're that blunt. Oh, they're so blunt. The Ukrainians, they are so blunt. They're just like, this fucking sucks. Here's what you need to change. Yes. Have you gotten any feedback where you heard like this fucking sucks and then you were were like no actually we should keep that that's the initial reaction every time pretty much which is which is like classic you know you tell an engineer they're engineering shit and then they're like uh yeah exactly um and you know sometimes sometimes we're right usually we're wrong um and you know you like i said it goes back to end user empathy like okay the end user does not understand the electrical engineering behind us nor do they understand you know deeply how the system works at a technical level, but they understand how to employ it.

50:49And the engineers usually don't understand how to employ it. So it goes back to a respect for the end users. And Henry Ford said, if I had asked my customers what they wanted, they would have told me, build a faster horse. And I think that's really important. We don't ask the end users, what power amplifier should we switch to? How do we solve this problem? But we do hear that this radio is too weak, and this radio is getting jammed. Not this one, but the one in Charlie. And then we came back with with this version and you know the radio performs 10 times better or something like that is that one of those things where when you get that you know that memo back or whatever can you rapidly incorporate that or is that the six month late time where you have to be shipping something that you know is suboptimal for months in order to just keep on shipping product while you basically reincorporate uh whatever that feedback is every day you wait war fighters die yeah and that's the that's the important imagine that's a massive fire under you yeah yeah and like you know we uh for the first many many many drones uh you know our signal contact was in the qr code in the in the box so it's like not just like just text us warfighter like some guy calls you and like you have cell service in the trenches and it's like not working uh like this is wrong um and you can imagine that creates a lot of pressure to make that six months less um and a lot of times it's less a lot of times it has to be more you just cannot build an artificial intelligence based drone swarm in a six-month product design cycle it doesn't work um and you know the engineer in me wants a very very long product design cycle but i love having the pressure from the end users from the um procurement as well in ukraine especially, that innovation has to happen very, very quickly.

52:42And we have to get really, really creative about how we build our supply chain, build our manufacturing strategy around product flexibility. How many of those calls have you actually gotten? Has there been a call where you're sleeping or something and someone calls you in the middle of the night and says, you know, things not working? Yeah. I need this fixed. and then you were able to like remotely support them and fix it in order to like complete some mission yeah and then you get the video back wow um and it's it's definitely not good for uh one's sleep schedule no no but or or sanity maybe but it's important it's important and it's like we're not building abstract things we're not like building like some satellite that you know might put a dot on someone and or i don't know like we realize that we build drones that put warheads on foreheads that that's what they do and we don't we don't hide that and i uh i live for it like knowing the uh actual impact of the product um and and uh seeing seeing as close as i can get without you know uh actually being like on the battlefield with him necessarily yeah yeah i i think both you and soren you talked about this a little bit earlier but you guys are flying to meet these people you know whoever your end users are in person talking with them building those relationships um do you have any good really good stories of you know some inflection point from meeting some person because you showed up in person there is there is no replacement for face-to-face time with end users and it is such a fun part of my job uh i think it's actually kind of like what i count as vacation um like i i i'll like you know stay here on a crazy technical sprint and I'll be beat.

54:30Like it just drains you. And then I'll get to go hang out with end users with the product. And I'm like, yes, that is so fun. I'm so refreshed. I can go, go grind so much harder. But there's definitely been times, especially with US end users that like, you know, you go to these live exercises, you go to, uh, you know, certain, certain events that I probably can't give too many details on and hang out with these people that are just such incredible people, such masters of their craft, especially since the people that are using FPV, uh, or at least historically that have been using FTV in the U.S.

55:01forces are really the tier ones, the most incredible warriors that America has produced, the people that are the world champion level of being a warrior and getting to see how they operate, getting to see their tactics, getting to see how they think, getting to see what questions do they ask and what goes wrong, and just standing there and watching them, not telling them how to do it, just watching them see like how do you how do you take out of the box how do you you know do you read the manual or not do you um and and building the product around that critical so so so critical um and also just like Getting to hang out with them and understand them as a person helps make it that much more logical to have that obsession with end user empathy.

55:50Like these are the people, you know, it's, they're just like, they first dance, right. That are, that are using the product that we are building the product for. Yeah. Do you have a list in your mind of, you know, you talked about like end user, like empathy for their experience and everything like that. Pure obsession over the line and how everything's manufactured and making the best product as fast as possible, shortest iteration cycle, that kind of stuff. What's kind of like the top things in your mind day to day? Uh, yeah, I'm, I'm pretty. uh, development timeline obsessed. Uh, it's definitely not like sexy.

56:24Um, yeah, I was like, I was like, uh, uh, random aside, but I was talking to a guy that like cliff jumps and he was like, you know, just like never, never, uh, you know, you, you never feel more alive, uh, than like the closer you are to death. Um, and then like, uh, you know, like Soren's like, he comes up and he's like, yeah, I love that's why you spend your day in notion timelines. like true true but hey that's that's what makes uh the that's what makes the product work so like top top of my mind day-to-day is like will you know next-gen radio ship on time will the design be qualified fully and released to production um when x engineer says they're going to get it done in this amount of time like is that for real and you know obviously you have to have a ton of trust But like, God damn, there's a lot of, a lot of, you know, importance on that assessment, on that, on that timeline.

57:22So I'd say like, you know, I go to bed thinking still very like tactically on like, is board release happening Monday for real, for real? Like actually, actually, actually. So that's, that's definitely top of mind, like day to day. you're trying you're basically trying to take whatever timeline people have or you guys have and push it up as much as you possibly can and try to also like like really think about is the timeline real yes is the timeline real and it's not only pushing it up often it's pushing it back really um in that like if you ship an unqualified design if you ship a design that you don't know works properly and then you build a thousand of it that is a way way worse problem to have than telling them, listen, sorry, it's not going to happen this week.

58:10Like it's just, I'm sorry, the physics are such that it will not happen this week. Production cannot start this week and you will have to wait until you start. That is a way better solution to be in than having a thousand drones sitting there that don't work that now the engineers need to go fix. And we've been there. The production hell of bringing these two systems up, we were like four months late, which is a long long time and you know we were like we were i mean i remember just like you know like nodding off in that in that corner like building these things um because we can't hand it off to production because we're still unsure on part of the design and that that's like that's been very uh uh informative like playing that that that game of like do you delay production or do you just go for it?

59:01And slow is fast in engineering, right? You win by doing design reviews early. You win by, I don't know, I'm pretty board obsessed. You win by spinning test boards before you integrate everything. You win by using dev boards. You win by quickly fabricating mechanical prototypes on our CNC back there and running flight tests. And sometimes it's kind of surprising how people don't want to run flight tests, but we are flight test obsessed. Why don't people want to run flight tests? Um, there's good reason, like, you know, especially when you're developing radios, like you're doing basically fundamental physics.

59:37You know, if you put your hand here just near this antenna, everything changes. The entire system changes. If you, you know, turn your phone on and pick up a call, that interferes. If your Starlink is going on, it interferes. It's a very, like, hypersensitive field. A finicky thing. A finicky field. So you want to do everything, you know, conductively wired in an anechoic chamber with isolation, blah, blah, blah, blah, blah, blah. and you know when you get back from flight test drone no work drone crash uh radio breaky like how do you debug that you're like well 10 000 things could have gone wrong so there's a lot of reasons like not to flight test um with that being said i'm flight test obsessed and we've got an incredible team led by levi the fpb specialist team and they go and they rip the system every freaking day and we have a we have a beyond visual line of sight waiver for manhattan beach so we are you know are you like the only company with that um i'm sure there are others could get it uh i've never seen another company there um but you know being able to um operate our our drones near our facility hugely valuable and being able to get reps on the system regardless of even if like some of it's unscientific but knowing like you know i i get i should get confidence on the design like ideally through you know lab tests but i don't feel good about shipping the design until we have hundreds of flights and uh that's like the kind of like that's entirely entirely reasonable it's like it's like shipping starship without actually you know putting people on starship and just hoping for the best for sure in version one and a lot of great engineering like they don't you know they don't test like they are confident that the design will work in space before they put it in space right and they do a lot of testing in the lab to build that but we are very like flight test obsessed and you want contact with reality early and often exactly and sometimes like i don't know what i do day to day it's a lot of like like i said timelining project management but also just like knowing when to push.

1:01:28Like, sorry, you don't get to do your scientific test. And I know you're not going to be able to root cause it thoroughly if you just go fly it, but it is time to go freaking fly it. And that's across mechanical designs. It's across electrical designs. It's across firmware. It's across autonomy. It's across the full stack of engineering. Crazy. So when you're under the gun, you have four months where you want to be shipping something, but you just can't. And I imagine there's like insane pressure every day that you're not doing this. basically people are probably dying or some version of that like extreme level of pressure um how do you kind of cope with that and like under you know keeps like stay sane when you know that you've got like the if you ship one day earlier and you just make it work one day earlier maybe someone lives yeah yeah i don't think you can have like balance or be what other people would consider stay sane i remember the push we made for that 6 000 unit drone order when we developed a ground station for the first time like i fell asleep in that corner there we were working on a bunch of fiber optic and i like found like a fiber optic spool and i just like just laid out and then like woke up and like everyone was like messing with me and shit and then like you know it was it was brutal um and we had you know definitely like people were getting sick and then they'd like you know and sometimes like people would uh you know work through their sickness which i don't know how is even possible um and we learned a lot in in pace um you know i think we would not have delivered on that demo in latvia if we hadn't had that that push um but we also learned that like that's not how you build a company that survives over 10 years um like i i definitely like would have physically broken down like i think i i would not uh be able to like you know physically do those kinds of hours for for more than than how many hours a week where you like was it one of those situations where you're basically elon in 20 we were sleeping in the factory sitting here there's bed bed over there so wake up work you know 18 hours or whatever and then fall asleep when you can't really function anymore and then wake up and do it again for four months in a row exactly yeah the the idcc sprint will live in infamy um and like it's it's cool to be in a you know in an industry where you have these like sprints it's like and then the entire team goes to lafayette together and we fucking kill it um we we the that we don't lose a single packet we have 100 link quality flying over the jammer um what was that feeling when that happened you've you've spent four months just and you have to act cool like in front of the the like big wigs and you just want to like jump up you want to jump up and you got like peter in the van our rf engineer mastermind behind a lot of the system you're with a spectrum and i'm like he's like losing his shit um and he's like but we're trying to no one can see him yeah yeah yeah and it's like uh the the like you know levi under the goggles saying cucumber cool like sticks you know you can't shake right and he's like you know our drone went eight kilometers before that idcc push and you know we we hit 10 we hit 15 um and then we you know we come back and we just keep hovering there and it's like five minutes of extra hover past that maximum maximum distance and it's like seven minutes which is like it just keeps going and going and going and the system just keeps working and just like acting like it's nonchalant and then seeing you know seeing how we did compared to the competition which is you know part of it seeing that we had the 100 meter cable working uh the very important to have the antenna very far from the end user and it was very hard to get you know power transmission analog video transmission over that that length and seeing that we had you know in-flight frequency changing for the video link working and seeing that our dual band link was doing what it what it should with with anti-jam capabilities and seeing that you're able to identify the target at range and just like all these things going right over and over again and then you get like you know of course the the military procurement people are like they're you know very like you know kind of you know manly manly figures that are like oh yeah cool like a yeah we've had that for a while but it's good to see it you know it's good to see that it actually works and then being able to as a team like let loose like we we uh we celebrated very very very aggressively in in latvia um i got to imagine this felt like you know the fourth starship or the fourth you know initial i think was it like a falcon the first the fourth rocket that actually worked for spacex after the first three blew up yeah and like the fourth was the last pallet of cash that they could send into space hopefully that thing doesn't blow up and it did yes yes and you know legendary companies are built on these like these events and like painful experiences yeah on on that and like you know levi and peter and glab and soren and the the squad from latvia like that that night it really cemented like i think like lifelong partnerships um and you know i'm not gonna say that the you know i i uh i loved every moment leading up to that event.

1:06:55But man, I look back on it so fondly because of the success. And I think part of it is like, you know, you can keep sprinting as long as you're winning. What really, really sucks is sprinting and losing. You know, those three launches before things worked for SpaceX, I'm sure. That was years. Yeah. And I'm sure people were lost, right? You know, some people don't have it in them to fight that hard and then lose over and over again. What do you think you learned about yourself and like what what did you learn about the team during those like four or five month period? I think it reminded me that like.

1:07:30Happiness is not about. Any given state, like you cannot perpetually be happy, you can only be happy based on. Becoming less unhappy, if that makes sense. There's no like stability. there's struggle and then joy of the struggle ending kind of you can be in the pit and you see yourself climbing out of the pit one step at a time and that like it doesn't necessarily provide happiness but it's like satisfaction and you know joy of like we're getting closer to the the promised land kind of thing yes and i think that like i don't know uh i'm really young and i haven't really figured this out but i see myself just like like i don't i don't really see like an end like what uh you know if we make a million drones i'm gonna be very very happy because that was a lot of pit but then i'm going to be unhappy and it's going to be 10 million drones and i think it'll keep going um i also i think you have to learn like that's the that's like the the devil on the shoulder right the like we're going to keep pushing for fucking ever which is very true um i just think of it as perpetually unsatisfied yes yes exactly and obsessed with making it better there's also zooming out and being strategic and realizing that like sprints like that are not how you build like decade-long companies um and not how you you don't win by heroics you win by careful planning just execution day after day over time exactly and you know you don't build a million drones on a four-month sprint.

1:09:08You build it by betting that the DoD is going to buy a lot or whoever's your customer is going to buy a lot and being really smart in your supply chain and thinking long, thinking strategically. And I think it was a formative experience. And I learned a lot about company building and what Nero should be about balancing those two very opposite requirements of let's get this tech working, let's get this prototype out, let's make the system perform, but let's also think on long development timelines. Let's not work 16-hour days. Let's work 12-hour days, or who knows, maybe an eight-hour day eventually.

1:09:48And let's build a company that people can live at. And Peter had a daughter, a very young daughter that he wants to be a good father too as an example um and uh like he has done both raising his daughter and building your srf um together which i respect and you know i i remember there's a you know a great image of uh alaya his daughter in in his arms and he's soldering with with alaya in his arms and he's like yeah she's gonna learn how to solder very very young um and i think that's awesome i think that like there there is there is else in life besides uh building uh drones at at high volume um but not much not much okay so when you're thinking about like massively scaling up uh drone production mainly that has to do with like figuring out how to make this factory way more efficient and then also scaling to more factories um how do you think about like designing the machine that builds the machine?

1:10:49That's a good question. There's a lot of good guidance on building factories out there on the internet, right? You listen to Ben Rich and you put the engineers next to the production line. It's very obvious. You have manufacturing engineers in design reviews. You can probably ask ChatGPT and it will output all this stuff. um i think maybe what's not as known is how unmanufacturable legacy defense is um what do you mean by that one of my uh my mentors very very very good engineer um ben stabler uh he told me something that i found pretty funny that's very very true um whenever you find like a part in electrical engineering he's also an electrical engineer that is mil-spec or defense, defense product, an aerospace product.

1:11:51It generally just means it was designed by someone who was fairly intelligent in 1970 and then never changed. That's very true. As an example, this part of our product right here is one of the only made-for-defense components that we use, and it's one of the worst parts of the entire system. um that that connector is depending on where you buy it over 200 that that little connector right there is so dumb so so so dumb and it's an embarrassment that that that is on a neuras product and it's being removed as soon as possible but the like that that connector is also like a standard issue dod connector it's part of um like the net warrior system i'll spare you the details but that connector is used on like so many freaking defense products it's on the it's on the abrams tank um and in my opinion there is a like a mind virus in unmanufacturable parts in defense and just being in defense pushes you towards these parts and like we're pretty i think we're about as forward thinking as you can get and like we made toys we made ftd racing toys that were so freaking cheap 500 drones that go zero to 100 miles an hour in a second or so um like we we are the opposite of aerospace grade designed in 1970 and never changed and yet mighty mouse connector on that thing um so it's like and it's not it's not like you know like i i woke up one day and found a mighty mouse connector on my product like uh it was a decision that we had to make it was it was forced but it was the same thing with like using china made parts at the beginning because that's the only like that's the best thing to do at the moment exactly exactly and so part of part of what I think about like part of what we do with being able to mass manufacture stuff is don't take a defense product and then try and mass manufacture it.

1:13:45Or like take a defense only an aerospace part and try and get it cheaply. Don't take a massive complicated FPGA and squeeze it into a low cost system. Take a consumer system. Take a toy drone. Take a off the shelf consumer component and upgrade it. Take an RF chip that's used for parking meters and manipulate its drivers and manipulate its different channels for different countries, different frequencies for different countries to make it into an anti-jam system that's very, very wideband. That's actually how we implemented our wideband anti-jam radio. Take a STM32F411, a$3 MCU that runs at 48 megahertz on a Cortex-M0+.

1:14:37These are ridiculously weak numbers. And put it in charge of the entire drone. Have it be the most powerful piece of silicon in the entire drone. And take a... It's cheap stuff all around. And cheap isn't the right word. It's inexpensive stuff, really. It's the same situation, I think, with Elon and the rocket where you could buy the$500 ,000 to a million dollar AC unit to throw on the rocket. Or you can just retrofit the one from the refrigerator that costs$150. Economies of scale. It's crazy. Economics 101. When you buy consumer products in volume, they become so much cheaper. And so we do a mix.

1:15:20We follow the advice. it's not obvious, but it's tried and true. The Elon advice, you know, put the production line right next to the engineers, the DFM obsession, the like, no, you can't freaking have direct solders on your drone. You need to have everything connectorized. No, you cannot have five different boards in the product. You know, three-board drone should be a one-board drone eventually. But then also we fight back against the legacy defense component mind virus of ultra expensive and not that performant parts that just never change. Do you guys have something similar to Elon's, the algorithm?

1:16:00Yeah, yeah, yeah. And Sean, our COO, spent 12 years at SpaceX and he brings the SpaceX wisdom there and he wrote up the algorithm and made us all memorize it. Oh, do you guys literally copy it? Yeah, I mean, it's in our notion. You know, some things you don't change. And, you know, the responsible engineering mentality, that's great. And there's a lot of problems with SpaceX culture. There's a lot of great things about it. And we're trying to copy as much of what works as possible. Yeah, exactly. Exactly. And it's been super helpful. You know, like SpaceX, it's like, I think, you know, already, but like historians are going to look back on like LA hard tech was defined by SpaceX, really.

1:16:38Like, it's like, you know, like all these startups are, a lot of them are founded by, you know, SpaceX, people that leave SpaceX and understand what SpaceX has done because What they've done is just never happened in human history, really. So, yeah, it's important to pull the right things. They do make a lot of mistakes. And, you know, getting to see the Bastrop PCB, again, I'm a PCB nerd, seeing the PCB production line in Bastrop, Texas, where they're making 5 million Starlink user terminals per year. Just wild. No one in America, besides ultra-low-volume defense primes, has built a PCB line, not the PCBA, not the pick-and-place assembly process, but the bare board.

1:17:25No one has built a vertically integrated production line like SpaceX has done. And they've done it so, so effectively. They are doing the hardcore chemical material science, like lithography, where you take sheets of copper and you pour, essentially I'm simplifying, something like an acid that removes copper to make these microscopic patterns that connect all these chips, that build the city that powers these drones. They're doing that all in-house and they have, you know, that's of different chemicals and they're bringing in these raw materials, these raw fibers and building it into these ultra refined pieces of electronics all in-house because the industrial base doesn't exist in America.

1:18:10And just building that capacity is so incredible. And of course, getting to see Starbase, seeing the city that they are erecting out of the middle of fucking nowhere to produce a really, really humanity-defining product. And then building a city. Really, they're building a city. Seeing that is inspiring. And that's where Nearest needs to go. We need to have—this is a detail, but it presents the picture. We need to have a board manufacturing division that makes millions of boards in America because there just doesn't exist the industrial base. In China, that's no big deal. You have JLCPCB that rips out these things like no problem.

1:18:50But we need to solve these hard problems. And obviously, it expands well, well, well beyond boards. It goes into camera optics. It goes into the ground systems. It goes into the antennas. It goes into the injection molding of propellers, the injection molding of the frame eventually. but we need to build that core capability that production capability here like spacex has am i am i right that spacex uh with starlink built basically like the highest um or the largest production american-made manufactured product like out there or complex engineering you know product and i'd imagine that like neuros is going to be kind of the second you know an iteration of Starlink, instead of putting out these terminals, you're going to be putting out drones.

1:19:35But it'll all be American. I'll zoom out a little bit. I think that there have been ages in human history, like the Steel Age, the Bronze Age. And I think we live in the consumer electronics age, right? And I don't think anyone in America does consumer electronics except for SpaceX well. um so i think absolutely to answer your question um and you try and get a board made in america you try and get this flight controller pick and place in america they'll charge you 120 dollars i've quoted it you try and do that at jlc pcb in china at low volume and they'll do it for under two dollars the delta there is 60x yeah it's it's bonkers and no wonder you can't make consumer electronics in America.

1:20:23No one is doing it except SpaceX. And the way they've done it is so hard. It's just like, inherit all the problems possible. You know, in China, incredible country. I went to Shenzhen and we've got some serious competition. You can wake up and order a board at 9 a.m. and they will etch away, they will create this multi-layer microscopic, you know, a thousandth of an inch precise copper pattern to connect all these chips, weld the chips, solder the chips on there, test the system in one morning. You can order at 9 a.m. and get your board at lunch. In America, it's like, I'm fighting like hell for one week board spins.

1:21:06I think that's key to our company. Right now we're at like three week board spins on certain items. It's so much slower. And SpaceX developed this product without that industrial base, right? And then developed the production capability without that industrial base. Also like without the tribal knowledge. without the tribal knowledge without the talent pool without the work ethic that a lot of you know obviously people at space like works very hard but like a lot of the attacks in china you just can't come in 96 or like hardcore and just like the commitment to quality the fact that they're okay living at the factory that's like standard uh in segregated dorms i mean it's rough um but um yeah just the the baseline like insane work the migrant workforce the fact that the government is willing to just ship you know laborers from certain provinces into the cities just to create to lower wages and then ship them back out like the government exists to kind of allow for consumer electronics manufacturing i think to a certain degree um spacex did did the impossible all without it and not because they were regulated into like they could make star link in china they just think that it's strategically a bad idea and it's the wrong thing to do for the country and elon just fucking did it um and that is that is awesome and their boards are competitive with China.

1:22:17They are making Starlink's for cheap, right? You can buy a Starlink. It's like$250 or something. Yeah, and I think I'd get in a lot of trouble for, and probably don't know the real numbers, but for sharing the numbers I saw on the screen on how cheap their stuff is. But I believe - They're making it at very low cost. They are making it for China prices, and no one does that. No one does that. So, yeah. Wow. I'm sure you can tell that this is where Nearest needs to go. Yeah. Okay. So when you're trying to hire people, this is a very... You're trying to build a team of A-plus killers that are just absolutely all in on the mission.

1:22:57How do you find those types of folks? And what's the process for onboarding them? They're never in plain sight. Soren found each other on Discord. And we would hang out 24-7. You know, I like in middle high school, like I didn't have like normal friends. I had, I had discord and I got to hang out with my friends, you know, for two weeks in July in Vermont a year. Um, so you find them in weird places. And like one of my first hires I met through Reddit, um, a lot of personal connections. Um, and also a lot of like, you know, being good at that initial screen. Um, super, super, super important. Like one of my best engineers had his resume was terrible.

1:23:44um it was like his question was bad and i was like convinced into taking the screen in 10 minutes i you were like this guy's amazing yeah yeah and it's like it's uh less scientific than i'd like it to be and it's something that like getting wrong kills the company because you know one bad player really drags down the average um but i'd say how do you build a good team let's look at like asymmetric recruiting opportunities like look at look at the discord servers look at the reddits ted does this maybe better than anyone i know put the sticker on uh the tesla charger that i see every time or the billboard next to spacex um and be really really good at that initial screen um and then grow them right like have a quantitative process uh have fixed questions that change every so often that way you don't have them memorized but like judge the different candidates on the same set of questions and put them through the technical ringer, do a case study.

1:24:47You know, I pull up a schematic with every electrical engineer I talk to and still like we haven't hired single electrical engineer that I haven't personally grilled at the schematic level. And that's awesome. And I love it. And you can tell so much about whether someone knows their stuff or not just by giving them a, giving them a like somewhat concrete, like technical, you know, challenge and then just working through it together. seeing how their minds work exactly exactly while still maintaining similarity like i've been i've kind of like sci-op myself before where i think i didn't make the hire but like i've felt so highly about candidate and then like finally gave them a quantitative like here's the same schematic everyone else got and they fail it and i was like oh shit wow that's scary like i almost hired that guy um have you had any situations where you just went out on like one of these missions or you go find someone there we're like holy shit we gotta have this guy yeah um it's funny last actually like pavlo who runs our ukrainian entity our chief engineer in ukraine uh definitely one of those wow um what was that story that was incredible it's a trip but we got introduced through an investor uh through d3 very very helpful group of individuals um and we just started talking and we're like we were talking about video systems and we're like he's really into like dvbt based digital video systems and i was like dude i'm i'm big on analog the union directionality and you know if you're doing like forward error correction you're just never never going to get the same performance blah blah blah blah blah can't have bi-directional communication just going so deep into like the technical debate and it just lasted hours and i was there like it was like me pablo and and uh someone who's on our team uh doing mostly on the vd side and just like watching uh you know I love him, of course, but watching his eyes just glaze over.

1:26:34He's like, these guys are never going to stop freaking talking. That level of just shared obsession. And we've all read the same Reddit posts, and we're all in the same weird-ass forums, and we're all reading the same - You're making some esoteric reference, and they already know it. Exactly, exactly. And that's the kind of person that you want to - Hire immediately. Hire immediately, but also give resources to and put in charge. And Pablo has done great things. And it's so cool to have, you know, engineering in country. And candidly, you know, this drone is the best in the world in a lot of ways.

1:27:09It's also very behind the state of the art in Ukraine in other ways, based on the fact that they innovate faster. And, you know, I had a really, really impactful lesson in Zaporizhia. Within Glidebomb range, they had an SMT line, a production line. that picks up chips, places it on boards underground where they were fabricating radios, you know, kilometers from the front line. So there was like, you know, dude with a cigarette. Like it's not very scientific here. Straight out of Breaking Bad. Yeah, well, but Jesse was running the show, not Walt. And like, you know, they'd send the guy to the front line.

1:27:54They'd take a spectrum measurement and they'd come back to the lab like that night and then populate a new matching network, put the resistor, capacitors, inductors on the SMT line, and then place the components for a specific set of frequencies that Russians are not on that day and then fly the drones next day. Wow. That level, like, we don't do that here, right? No. It's impossible. Like, just the flight. The flight from here and the train, sum those two together, no engineering time, zero minutes, is longer than the feedback loop to product iteration in Ukraine. So, obviously, we have to have the respect to know that, like, this drone is going to be behind in in some ways but ways yes yeah but so having the only way to really catch up is having engineering in country and give them the innovation potential and i think pablo has done an incredible job at that crazy okay so like long term when you're thinking about entering some country or whatever you're going to be shipping a lot of product let's say to ukraine for now and other places you know soon um are you thinking about just spinning up an entire factory in that country so that you can do like local iteration really fast yeah Yeah, it's, again, there's a lot of dreams, and we can't let any of our dreams get in the way of our core dream, which is million dreads a year here in America.

1:29:08That's the number one priority. With that being said, you have to have localized logistics. You have to have production. And as a country, if I was in charge of defense procurement, I would only buy from in my country. I would only invest in local production capability and local jobs and in the local ecosystem. And I think Niros, in time, can do a very good job of bringing local production capacity to countries, having entities in those countries. And as an example, our Ukrainian entity would be awesome if they built drones in Ukraine. That would be so freaking cool. Maybe in Taiwan, we could have an entity, a Taiwanese entity, that is run by Taiwanese people, that is building the Taiwanese industrial base, and that's building Taiwanese production capability.

1:29:56And I reject the idea of how Apple develops products where they just have a contract manufacturer. We're too good to do the manufacturing. I'll outsource that manufacturing. This isn't what we're doing. It's not like we're hiring a Taiwanese CM. We are giving the Taiwanese entity or the Taiwanese effort the ability to generate production internal to NEROS. um and i think that's that's really important and you wanted to copy more of the tesla model where they actually build their own you know non-partnered factory in china and it obviously employs chinese people but it's a neuros factory exactly we're not hiring foxconn like apple is not to say i hate apple but we're we're we're not hiring foxconn like apple is we're building gigafactory shenzhen like tesla is awesome okay let's uh let's end it on what's the hardest thing you've overcome?

1:30:48Hardest thing I've overcome. Um, I'd say I've lived a lot of my life very split. Um, like, you know, in, in middle school, I wouldn't really talk to too many people and I'd like just like spend all day dreaming about like, when can I get home and fly? Um, and I remember just like, just like moving my thumbs in class and just like so obsessed with like doing something that I'm not doing. And I think that like, you know, obviously I love that, but definitely it took a little bit of a toll. Like I think I am a, like, I love interacting with people. People are the most incredible machines imaginable.

1:31:29And it took a toll on me to be so disconnected from, from people. And so obsessed with the online world, like all my friends, you know, Soren and and Liam, and Will, and all the, you know, Nick, the group of people were all just Discord emojis, kind of, or Discord profile photos. And I'd say that, like, living that split life of just, like, and trying to do well in school, and, you know, like, I was pretty academically focused as well. I, you know, was very lucky to go to a good public school in the DC area, and, you know, my parents really, really, really wanted me to do well, and I cared a lot about my SAT score and things like that um you felt like much more connected to the people that you were communicating with like over discord yeah i mean i didn't have friends in high school based on like like i just went to school to like get a good grade maintain my gpa and then can i please go home and fly um and just like get this over with as quickly as possible and i definitely feel like uh went through a like not so great time um where i was just like so fed up with only online friends and wanting like in-person friends.

1:32:38And I think like that was, you know, started to solve by leaving college and going to hang out with people that do startups because like building was really, like at the core of drones, it's building. And then really felt fulfilled via Neuros because my friends are these people that build the front. You're finally getting to kind of be in the environment where you don't have to go home. you're just here you get to be here all the time I'm just here and there's no discord server there is my friends here