The Exchanges

Q And you were, and what were you doing with circuits before you were doing EPRS?

A Yeah. So, uh, so at Georgia Tech, we were building, um, silicon neurons. So the, uh, thesis was you can, we worked on the DARPA, uh, synapse program where the idea is, um, if, if you build, uh, neurons out of, um, silicon that work much more like the brain, you can operate much more efficiently than like what you can on a GPU. So then from there went to, um, um, IBM, uh, worked on Watson. So I was on the, on the chip team that built Uh, the chips were Watson. That, that was just a great experience. I ended up, uh, there. And then I've always thought, um, and I think, uh, very similar to, uh, your, uh, some of your thoughts, Joe, that I share like this strong passion for defense. So I wanted to help out. And so I went to, to Raytheon, uh, to work in the aerospace and defense.

AI assessment note: “at Georgia Tech, we were building, um, silicon neurons.”

Q Why, why, why hasn't this been done a long time ago? Is it, would this have been a lot harder to do with chips? 20 years ago.

A Yeah, so the big thing that we've discovered is, it's like, uh, I was, it's like, uh, uh, uh, the power line example. So like, if, if I want to manage, uh, like a power, the power going through a power line, right now, uh, outside that powers all of our houses, there's this giant transformer that uses 20, uh, kilovolt lines. Um, previously there was no hardware to manage, uh, you can't, You know, a digitize a 20, a kilovolt line. There's no hardware that goes high enough power. And so the same, the same analogies on a vehicle, right? 600 volt is a common voltage. It's a flowing around a lot of these vehicles. 800 volt in some of the new Porsches electric that are coming out. And so previously there was no hardware. I couldn't, it was all analog. I couldn't digitize it because that's too high of a voltage. So what we've invented is a gallium nitride chips and a silicon carbide chips. These new chips That can handle 1200 volts, 2000 volts. So we've digitized.

AI assessment note: “previously there was no hardware... we've invented is a gallium nitride chips”

Q So you have all this power flowing through this system and you want it to all flow to hit at once in a way that's efficient. And so, and so the chip's fast enough to kind of talk to all these different parts of the system and coordinate the power or what's it doing?

A Exactly. That's exactly right, Joe. So the, actually the, um, analogy I like to use is, uh, I, so I live in a Manhattan beach, uh, California down in Los Angeles. Um, my wife makes me water the garden. So it's like the, uh, the analogy I like to use is like, if you have a old school hose, you know, you just turn it on, you walk back to the trees, you're wasting just tons of water as you walk back in between plant to plant. Uh, what if you had a machine learning, uh, algorithm and hardware that you could hard, uh, cut off the hose. So there's no leakage whatsoever. And then right before you get to the.

AI assessment note: “Exactly. That's exactly right, Joe.”

Q So in building EPRESS, you're having to attract the best and brightest from the defense world, from different parts of the technology world. What's the culture like internally? And what are you trying to focus on there and for the culture?

A Yeah, I think, um, it's really simple for us. We focus on people first, uh, and it's, it's really paid off for us. I have employees that come up to me, um, every single day almost and say, Hey Bo, this is the greatest company I've ever worked for. And a lot of it is, uh, a couple of things. We challenge our employees to work on the most cutting edge of technology. And so true engineers, you know, really love that, um, about it. And then, but also eliminating waste out of the process. Cause our customers don't want us to waste time. Taxpayers don't want us to waste their money and engineers and operations folks and employees don't want their time wasted either. They want to be working on cool stuff.

AI assessment note: “We focus on people first, uh, and it's, it's really paid off for us.”

Q So is it important to you that America is strong? You, you, you believe in defending America and, you know, and you know, a lot of people attack what our DOD does. Like overall, your optimistic is better for the world of America's.

A Yeah, super, super optimistic, Joe. I think this, uh, this saying I like to use is like, I remember I read this article once in the Wall Street Journal that was like, hey, all of our PhDs are going to, um, help people, uh, uh, click ads more, which is, is great. That serves a really important part of the economy, but like, let's get some of our PhDs and some of our, our best engineers working on, Uh, really hard problems too, like in the defense industry. And I, yeah, I just, I, and what a lot of people don't realize is the history of Silicon Valley really started with defense, right? Like a Bill Packard from HP, like he was the secretary of defense, uh, the internet, like a self-driving cars, a lot of that technology started in the defense industry. So I think it's, it's so important.

AI assessment note: “Yeah, super, super optimistic, Joe.”

Q So you could use this for theoretically data centers for, for other, I mean, did other things in defense you're using it for? What's something else in defense you could use it for that you could tell us?

A Oh yeah, sure. So, um, right now there are, if like, let's say I want to put new capability on a ship or a new capability on a, a, a, A tank or a striker vehicle. Well, it takes power. And a lot of times they're at, these are systems are at their power limit. So what smart power is able, so we're actually working on a couple of projects right now, both with a ship board applications and like, uh, what I'll call a military vehicle applications where there's two big efforts going on right now is number one to electrify those things. So they're like electric, like you can imagine like a electric tank, right? Or a hybrid tank, right? Uh, and so managing all the power more efficiently in that vehicle. The other thing is, even for still a diesel driven vehicles, all that energy gets generated in a generator, and then you, maybe you have a radar that's blasting at really high power, and then you've got a laser that's blasting at really high power, or a. I have lots of lasers on my vehicle. Yeah, so it's managing and it's charging up and storing that energy and using machine intelligence to figure out exactly when that needs to fire.

AI assessment note: “working on a couple of projects right now, both with a ship board applications”

Q it seemed like it dropped out. It seemed like maybe starting in like the eighties, nineties, Silicon Valley started focusing on other things. And, and then this, this generation seems to be pretty against working in defense. As you're building a defense company, there are a lot of people who are attacking you for that or who are not interested because it's defense or what's, what's the, what's the energy?

A Yeah, we've had a surprisingly positive response, uh, to what we're doing, and I think that was one of the goals of starting, um, um, um, EPRES, um, with you, Joe, is that we wanted to bring that, that youth back into the aerospace and defense industry, and that was one of the really important founding, uh, missions of EPRES, because I saw it at Raytheon, right, is like the number of where I was just before, uh, we founded EPRES was like It was so hard to get millennials and some of the younger generation, younger kids, but, and I think that's why it's so important. Like a lot of, a lot of the kids coming out of college now, they're like, oh, a startup that does aerospace tech, like that's so cool. So I think that's one of the things I'm really optimistic about. I think by starting companies like Epirus, I think we're starting to pull the generation back into these really hard problems, uh, space tech and defense tech and aerospace.

AI assessment note: “we've had a surprisingly positive response, uh, to what we're doing”

Q So this cone of radiation from Leonidas can shoot down a drone. How far, how far away can you shoot it down from?

A Uh, so, uh, we have, um, demonstrated, uh, beyond, uh, uh, a visible range. Like, um, a good story I'll tell you is we, when we did our, one of our first, uh, field tests, uh, field demos, we had about a 70, uh, customers come, uh, generals and so forth. Uh, it was a great event, and one of the, one of the biggest technical challenges we had to solve is we were taking down the drones, uh, so far away. I remember, Um, Lee, our CEO, he, he comes over to me and he wants to watch. I was like, Hey, uh, you're going to have to use these, uh, binoculars to see the drone go down. It's going to go down so far away. And he's like, Oh man, I still can't see it through the binoculars. So we got a telescope. So one of the biggest technical challenges we had to do the demo for our, our customers was we snaked out like a, you know, the longest fiber optic cable we could possibly find and put a camera, a setup way, way out in the field. And pipe the video back to, um, and then we pipe some video back through drones and so forth.

AI assessment note: “we have, um, demonstrated, uh, beyond, uh, uh, a visible range.”