Aviv Shapira & Tal Horesh, XTEND | theCUBE + NYSE Wired: Defense Tech

Palo Alto studio connection Silicon Valley and Wall Street. I'm John F co here with Dave Volante my co-host. Hello, I'm John Furry, host of the cube here in the cub's NYC studio. Of course, we have our Peloto studio connecting Silicon Valley to Wall Street. We have a twofer here. We have IPO news as well as defense tech content. one of our hottest series where you know physical AI is really moving into the defense tech making the battlefield safer and more efficient for everyone. We got A Shapiro co-founder CE of Extend. Congratulations today and Tales CFO of Extend Rang the bell this morning on an IPO raised some good capital. Gentlemen, thank you for coming in. We got two topics. I don't know where to start first. I want to go deep on the drones and the physical AI, but let's start with the the top news. Congratulations. You're a public company. >> Yeah. >> Do you feel different? feel clean, you feel ready to go. >> Very big day, you know. Uh we're standing here, sitting here, just the two of us, but there are about 20 200 people >> all over the world, you know, like very very happy right now. >> This has been an incredible journey, you know, eight years to get here. But, you know, they said this is this is just the start, right? This is just the beginning. >> Yeah. I mean, you got to be all the scrutiny. You have to produce results. >> Yeah. >> Tall, talk about the the funding. Obviously, IPO is a funding event for you guys. You feel good about the capital? Yeah, we we received around $100 million in capital in during signing and closing. Now it's for us it's just the beginning of the road. We we wanted the access to the public market to get access to capital and in order to establish all our strategy going forward. >> You know, a lot of people ask me about defense tech and why I'm so focused on it right now. One is demand for anything physical AI, but you know, the battlefield and the war fighting environment is can be safer. um and all environments where you have people in harm's way, you know, police, other first responders, but it's also a tech angle because a lot of commercialization is coming out of this tech. This isn't like defense tech, old school like, hey, build this for this military contract, although there's contracts involved, but the tech is very commercialized. You see physical AI and manufacturing, retail, financial services obviously, but it's AI, it's the AI version of tech. So share with us what you guys are doing from a tech perspective because this is a physical AI story but okay it applies to defense tech but we all know what comes out of these >> military andor defense related government contracts is that >> the benefits spill over and accelerate value. >> Yeah definitely. So uh yes we are a physical AI company and a lot of people are asking what does that mean? Are you a drone company? Are you a physical app company? What's the difference? Um so you know in essence as you said rightfully so there are millions of people protecting us and we have the right and the privilege to protect them. Uh so we are basically building you know the the robotic war force that deals with some of the world's most dangerous and complex missions >> you know boarding buildings underground tunnels you know these very very dangerous situations. Uh we managed to build uh an operating system we call it XOS. Uh you saw the video. um you know the the operating system basically allows an operator, a human to actually just say what he wants to do and autonomy does the rest. Okay. So it's it's the most natural way for humans and machines to actually interact to solve some of these very very dangerous uh missions. >> Yeah. And if you look at the video, you can see the humans in involved. They have the drones are deploying them. They're very intentional in the execution. Um there's a purpose for that. I have to ask because you know this is a technical challenge. um these called the tactical edge uh uh where there's not a lot of com support at the edge. You connectivity challenges, power challenges. Um you have a lot of things going on with the the role of the human in war fighting. It's the it's the it's the first on the battlefield, which by the way, we've learned in the US that friendly fire ends up causing a lot of deaths, too, which a whole another discussion. But in the re in the real world, the tactical edge is us humans walking around, right? You're starting to see the edge. So compare and contrast because that's where physical AI is going. So talk about this tactical edge. Drones are coming in for delivery. We're seeing that now go mainstream. I wouldn't would have thought that never would have happened when I saw Amazon did that for you know that will never happen in my lifetime. No definitely you know like now it's happening. >> So let's start first of all what what is what is physical AI? You know what what is that buzzword? So everyone is used to using AI for text for recipes for you know uh like guidance on on learning things. Physical AI is exactly that but with a body. It has now the arms, the legs, the wings, the the motors. It it can actually go to places and actually replace you know us in these very very dangerous situations again in our market in in uh in defense and security. Um now the the most interesting thing here is that uh you know the drone people talk about drones you know and robots and and the most interesting thing is is these are just uh vehicles to get AI to the edge. These are just like like a truck. It's like a it's just a way to get from A to B. Once you get to B, the application is what matters. >> I was having conversations with Nvidia just yesterday and I've been talking about it since Mobile World Congress. When you create intelligence in these AI factories, you have to distribute it. So there's a lot of networking involved. Um and also the application whether it's on the battlefield or in real life when you inject intelligence into things they change. So I guess my question to you guys is what's your reaction to that? because you certainly could create better product at the edge. What's the role of like say Nvidia is AIing the edge when someone says hey you're bringing intelligence to the edge obviously robotics kicks in but what does that actually mean what's different about that versus say hey I'm just going to stand up a tower put a bunch of compute servers roll out a rack maybe put some compute >> so you said two two important things you know uh edge edge uh you know edge devices and he also and he also mentioned the fact that in in wartime it's a contested environment you lose lose communication. Sometimes there's no more humans or operators connected to the machines. So these robots have to actually operate on their own. So [snorts] we work with amazing companies like Nvidia at the edge on the ground. Um and once you are disconnected from the operator, these machines have to actually act on their own and this is where AI kicks in. Everything of course ethical everything is you know within the norms and the boundaries that we humans actually dictate. But they operate on their own. They talk to one another. They make decisions you know they they they And we can talk about examples but they actually get some of the you know the most dangerous missions solved on their own instead of humans actually going and risking themselves and you know taking physical. >> So the old networking was hey hello response hand shaking they have to reason on their own. So say they lose comms. >> Yeah >> that's where the AI so you have to have models resonant you got to have the right hardware. Talk about the requirements. What does it take to do that? So we work with uh you know all all all all all all all all uh a scale of of computation devices you know all the way from like Raspberry Pi Qualcomm to Jetson at the edge. >> The more computation you have the more AI applications you can run >> you know we started by by introducing what we call AI assisted flight. So you don't fly the drone you you basically tell it what to do but that still requires an operator to tell the drone what to do. Then we move to tasking. >> What does that mean? So you can actually task, you know, inspect the room, pick that up, blow that door. H But now it gets interesting because now you can actually deliver an entire mission. You know, you can tell a bunch of drones, you know, you two go to that building, you two go to the to to fly above and you two robots, you know, enter the building and look for survivors. >> And that whole orchestration is can actually be done by AI and they're going to talk between themselves. >> So a lot of computer vision, >> a lot of processing power. All right. So everything's bounded by power and money in in the AI world. Put the money thing aside for a second. power drones require batteries or electric um some cases they want to be undetectable share your thoughts on the power requirements how you view that because you got to power the AI >> so for us it's much easier you know at the end of the day a drone is just a flying iPhone you know the but the electrical motors uh consume something like 100 times more power than the Nvidia Jetson so power is not the biggest war we have right now >> is it so you feel good about the power >> yeah I mean uh it's going going to take time until AI is going to consume more energy than the motor. >> How about the techability? >> The techability. So the it's actually interesting that the smaller the robot is or it's more the drone is the more noise it makes. >> Okay. Uh just because of the propellers, they have to spin much quicker and they're actually you know I'm an aerospace engineer so they actually surpass the speed of sound. >> So there's like a hyper uh you know hypersonic boom. Anyway, too much too much detail. >> I can I can tell those small drones they they buzz. >> They buzz. >> They buzz a lot. By the way, we got some feedback from the from the battlefield in Israel that uh this noise actually helped the the soldiers to to like uh to operate uh only you can detect the the drone and also the the other side was was afraid from the noise from time to time. >> So they run [laughter] that is actually humble. He's never going to just say it but uh he's a soft himself. is uh you know special operations and and working in extend. >> I wish I had drones back then back back in the days 20 years ago but uh now >> how much has it changed tactical control because he just laid out a orchestration scenario that quite frankly is impressive because you can go in and replace potential harm's way. I can say I can say with a lot of confidence that extent changed the battlefield in Israel because unfortunately we had a few conflicts in the last years and the drones was been able to be a flying camera and you can put a camera almost everywhere also inside inside indoors and outdoor. So imagine yourself that you even don't need the K1 K9 dogs to go in. You can fly the drone inside. So you can save lives and this is our mission today. >> All right. Talk about some of the momentum in the business. Obviously you got good fundraising. Congratulations on the IPO. Uh obviously it's a staple in in the on in the arena. Um >> congratulations. Talk about the production. Drone production is in demand. >> Um >> the old school methods of the big Patriot missiles. We heard about the Iron Dome and all the telemetry involved. There's so much tech, >> but drones changed the game. >> So there's a demand for drones. you guys pumping out a lot of drones. Can you share some numbers? >> Yeah, we we we are we see the market uh demand like worldwide. That's why our part of our go to market is to be global. We have today five locations that we are operating with subsidiaries. Uh in Israel, in Singapore, Latvia, UK and the US, Singapore, Latvia, it's acquisition that we did in the recent years. Uh also tech acquisition, but also market penetr penetration. And all of these five have local integration, local manufacturing and local sales and development. All the five using all the group assets in terms of platforms and in ter of capabilities and that's why we can we can scale uh much more than others. >> So any and compared to others are doing how would you rank on the volume side of it right now? So the by the end of this year we should have about 25,000 uh systems out there both deployed and contracted you know the the what Tal mentioned about uh we call them exps these these individual companies they actually help one another when war started uh in Israel you know everyone was helping Israel uh now that the US is ramping up industrial base you know we're doing a lot of technology transfer to the US building the industrial base here in the US >> they certainly need help here in the US I know from interviewing multiple drone manufacture They don't have the volume you guys have. >> Every country want self-reliance or not only from manufacturing perspective al also supply chain and technology. >> So is that a licensing deal for you guys or is that more of you guys come in and set up and build with them? >> Um so in in in most cases we build our own factories but some in some cases we have to outsource because it's like a sinus of demand. Uh we have a pretty large factory in Tampa. you know, we're part of a drone dominance program that requires, by the way, it's it's an amazing program that actually gets company like us to scale it production. >> Um, so, um, you know, for us, uh, we're now ramping up the, uh, you know, the Tampa facility, just doubling it >> to reach, uh, something like, uh, you know, what's the number like a thousand drones a month. >> So, on on the, um, investor side, what can people expect now that they're out there? Stocks out there. I don't make stock recommendations. I just share data although I do have my own favorites. Uh I I don't even I need a disclosure. I'm not a work for the NYC but um and investors what can they expect from you guys? Um what >> I mean I'll start sorry I mean the recent momentum you know we signed a $15 million contract to replace Chinese drones in Europe. It's a big deal. Taking that technology now to the US has you know infinite potential for uh the current gap in the market. uh we had some amazing deals in APAC uh Israel, US so you know the the market is you know the market is growing you see the budgets are growing and we're growing with it with it >> we we are great we are creating we are going after the large projects uh currently our pipeline is half a billion dollars it's qualified deals that we are going after we we been in charge touch touch with the decision makers and the budget allocated and uh that's what we are targeting the out product in each country. >> Yeah. Talk about the OS. I love and in your video there's a mention of there's an operating system controlling they talk to each other. I see a lot of commercial benefits because when AI factories grow to the edge and the metro environments, you're going to have basically distributed computing network. >> Yeah, >> that's going to be AI architecture. >> It's kind of the same but a little it's different because it's got AI in it pumping out tokens. They'll be powering drones. I could see someone saying, "Hey, you know, I'm at a telco endpoint node tower central office. Deploy some drones. check out the territory, check the the antennas, go check the power lines. We've seen this commercial use case, but those drones are huge. Drones are highly expensive. So, is there a cost structure of efficiency on the commercial side? Can you guys uh share? I'm sure it's increased your TAM in significantly. Is that something you're going after? >> You're probably uh uh hearing and talking about a lot of companies trying to build as many drones as possible. you know if if a country says to a certain company build me 10,000 drones I don't hear them saying build me 10,000 pilots right so the the bottleneck is now moving or it's going to move from the drones you know once we build the factories to the pilots and this is exactly what we do um about 7 years ago we understood that one operator has to control many drones infinite amount of drones and not from the field but from remote so once you have these you know millions of drones potentially we're going to be the operating system that that controls them you you know, just human intention. >> Yeah. >> Machine autonomy. That's the B >> and that's where the intelligence comes in you were saying earlier. >> So, we we've we've taught the system, you know, it's an operating system that is built on a skills layer. We taught it to do things. There there hundreds of skills, you know, go into the window, look for a people, do that, do this. And these skills, you can basically talk to the system or command the system. And this is what what it does. You don't pilot it. >> It's a prompt. >> It's a it's a it's a prompt. >> It's a prefill >> because [laughter] it's natural, right? decodes pre-filled and decode we're seeing disagregated serving becoming dominant. >> Exactly. And I I think you know um like like the you know the the old mobile days were 100 not yeah maybe hundreds of mobile companies and when iOS and Android came in they became the standard and everything was around there. So this is this is our ambition. >> You know, I think the whole operator piece because you know, we're in a market of build, operate, and invest. We're seeing that with in all the themes in AI, but the word operating becomes more and more prevalent as we get into this execution layer. And I think that I think the the cat's out of the bag, whatever you want to say about the fact that AI is real. You guys are proof of that. You're seeing the benefits on the product side, but now it's like how do I execute? >> Yeah. >> How do I operate this? How do I turn on not stand up? I mean, the language is changing even on my show. One one way, >> you know, stand up an infrastructure. No, no, no, just turn on. >> One one way to do it is create a lot of integration with a lot of platforms and other other components like you can do cameras, can do C2, you can do vehicles, you can do almost everything because once you have the the extend operating system inside when you integrate it with a lot of assets, you can scale. >> All right. Well, I have to talk about your journey because you have multiple ventures under your belt. Congratulations. >> Almost as much as you. [laughter] I'm not a I don't run a public company now. Um so um thanks for the compliment, but you guys are impressive. >> Entrepreneurship has been a big part of the culture and tech. You guys live it. >> Now that we're seeing AI and tech go mainstream, tech's in everything now. >> There's a lot more entrepreneurship happening right now. You're seeing a lot more energy around, hey, I can innovate and AI is certainly helping people. So I love what Nvidia is doing. I love what you guys are doing. What's the entrepreneurial opportunities that you guys see developing out of your business? And what would you say to someone watching um that um might have entrepreneur ideas? Is the road I mean certainly it's always hard to do a startup but it's not as hard as it was when you know we were grinding >> for a startup how hard it was. >> AI can make can create an amazing presentation for you the perfect pitch but executing is still something else. I mean executing is all about the people. Everyone who asked me you know what's like your battle biggest bottleneck you know what do you what what keeps you up at night? It's the people. So, uh, looking for the best people across the world is like, you know, the most important thing I would say, you know, you ask me to entrepreneurs, um, you know, this is my fourth company. I sold my last company to Intel. I have two Emmy awards. It doesn't matter. At the end of the day, you know, every every idea that you have, if it's a big idea, or small idea, people are going to tell you, "No, you're not. It's a >> terrible idea. Someone's doing that already." >> Yeah, Google can do it, Apple can do it, Amazon can do it, extend can do it, right? But, um, you know, that's that's the that's the magic. If you can listen to them, learn and go around. That's >> Yeah, that's great advice. I I appreciate that. You know, one thing you mentioned skills. I want to talk about this because I think we're probably living in the best time in my life that I've seen where if you're an engineer or a builder or creator or curious, you can do a lot of cool things. So, physical AI is attracting a unique skill set. people who love space, people who love air, aerospace, computer science, I mean everything, every kind of discipline. There's no one school you can go to in one degree, it seems like. I mean, who doesn't love drones? I mean, drones are cool. Like, they're flying around. It's physics. It's the real world. So, a lot of people, young and old, coming together as a fusion of talent that are attracted to physical air. How would you guys describe some of the constraints in the market? Because money follows the constraints, the bottlenecks. What are some of the challenges people could jump into or that might be available for folks? >> So, amazing software capabilities across the world. A 13-year-old kid can be the best programmer on the planet, right? Uh but where does he run his physical application on it? It cannot just buy a Best Buy drone, put his software and and go fly. And this is exactly what this is why exactly why uh we've built the operating system. So, we allow anyone in the world to go to the extend website, put his application, and hopefully it will get to the creating a programming interface into the drones to encourage developers >> and leverage your platform and hardware. >> We've created that. It's called the XOS marketplace. We have hundreds of applications there and uh some of these small applications from a 13-year-old kid in Brazil can reach the US government. >> So, one of the things you guys are doing that you mentioned is innovating. You've trained the drone in this case is a war battlefield scenarios. Can I train your drone if I'm a developer to go walk my dog? check the mail, find out what's going on with my kids. >> I'm just making that up. Those are examples of nonwar. But no, but this is my point about what creativity comes from. >> AI created amazing simulators as well. So, you know, you can actually get a like a hybrid or or a virtual simulator for drones and then you can practice any any algorithm. >> Guys, very intriguing. We would definitely love to have you back. Certainly, you come back next week. I want to see that drone. >> Um, I guess the final question is now that you got the money, you're public, what's the focus? What's the use of funds? What's the strategy? What's the what's the what's the next steps? What's going on? >> Generally speaking is to enhance our US entity because now we are fully on the extend AI robotics in Delaware. Uh we want to increase the human capital and uh also establish a strong supply chain in the US in order to deliver all the platforms as well as acquisitions strategic ones. uh but more to >> bottom line I mean yes that's that's >> inorganic acquisitions of the team >> these are the practical uh and like the actionable reasons but uh I can tell you that uh after saving many many lives all over the world you know getting our mission getting getting our technology out there to as many many uh end users as possible that's this is what the IPO is for >> yeah and also you it's critical people don't understand that the safety angle here is huge impact saving lives um prevent preventing loss is huge. Um, check on that. And I just love the the technology angle. Um, Nvidia's involved. You guys work with Nvidia. >> A quick mention about the relation with Nvidia. >> Um, not formally. >> No. Okay, cool. Um, thanks for coming on. Appreciate it. >> Thank you very much. >> Thank you. >> I'm John F with the cube. This defense tech and new hot IPO checking it out here. The physical AI continues to be a hot story. I think next year in 2027 we will see a lot more content coming out. Advances in robotics and software. More and more stuff coming here on the cube. Thanks for watching.