Down to our last two sessions, and they're two great ones. It's my absolute pleasure to welcome Don Burnette, Founder and Chief Executive Officer of Kodiak. It's been a pretty busy year since I think I last saw you live at CES. Kodiak Robotics, we have to say congratulations. It just actually went public via SPAC last week. Don's going to not only talk about the business, but obviously a lot that's happened in the recent month leading up to that. We're going to talk about AV trucking both in a structured and an unstructured environment. With that, Don, thank you so much. Great to see you again. Awesome. Thank you. Maybe I'll come over here. I'm going to walk through a couple of slides really quickly and then get to the Q&A, which I think will probably be more interesting for folks. It's great to be here. Thank you all for coming. I see a lot of familiar faces and folks that I hope to get to know better soon. Typical disclaimer about forward-looking statements that it's now obligatory as the fact that we've gone public, which is incredibly exciting. Kodiak Robotics has an AI-powered solution for autonomy. I know this is an autonomous conference, so you all are probably pretty familiar with that as well. We focus on the commercial sector. Starting out in long haul, we've moved into a couple of other different subsegments of commercial trucking as well that I'll talk about. Our commercial network, broadly speaking, is around 23,000 mi across the Southern United States. We were some great partners in the industry. Some of the folks who are here today. Thank you all for all the support along the way. We have active lanes that we operate on a daily basis. Those are the ones in yellow, mainly emanating out of Dallas, where we set up shop back in 2019. We've been operating freight out of our Dallas facility since then, Dallas to Houston, up to Oklahoma City, and out to Atlanta, which is one of our primary freight lanes. Since this is about autonomy, I wanted to take just a quick second and step back to talk about where we were and how we got here. It's been a long road. We started moving freight with our technology way back in 2019, as I mentioned. We did our first commercial delivery. We moved also into defense. We have a play in autonomy defense, which we may or may not talk about in the Q&A. We've brought on different makes and models of vehicles, including passenger vehicles like our Ford F-150. We launched a different generation, our sixth generation of the platform, which is our latest incarnation of the Kodiak Driver Solution. In December of 2024, we launched our first fully driverless product. That's a truck that's deployed on a daily basis for a paying customer with nobody in the cab. I'll get to a few more examples of that and some more detail. In 2025, it's been an incredibly exciting year as we continue to expand our deployment, gather more data, get more experience, especially from a driverless perspective under our belt, and really set up the future expansion for the company in years ahead. By the numbers, we've done about 7,300 loads. We've driven over 3 million mi with our autonomy solution. I mentioned our network size is 23,000 mi, and we have over 3,000 hours of paid driverless operations. We like to think about our driverless experience in terms of time. Time is generally how people value their worth. That's how drivers think about it. That's how people are employed. We think hours of paid operation, a driverless specifically paid operation, is one of the key metrics that folks should follow in the autonomy industry. A little bit about our focuses. We have both commercial trucking and our public sector business, where we work with the U.S. Army and a couple of other military armed forces. It's a very large TAM, as many of you will know. This is a huge opportunity from an autonomous perspective, both in the U.S. and globally. In recent years, we have focused in addition on unstructured environments. That's where our launch customer, Atlas Energy, comes from. We wanted to put together a perspective on the opportunity within that business. It's definitely smaller than the over-the-road trucking, but we think this is a massive market opportunity in our own right, covering areas like oil and gas, mining and mineral transportation, and then one that most people don't think about a lot, logging and timber transport. These are all typically unstructured environments in remote locations where you don't have easy access to infrastructure and roads, but you still need trucks to move goods, move materials, and transport things around. Today, that's only possible with drivers, human drivers in the seat. Tomorrow, we think all of this is going to be possible with an autonomous solution. We're generating driverless revenue today. We think this is a huge step in the journey for any autonomous company. We're very proud to say that we've now deployed eight driverless trucks that are operating 24/7, that's day in and day out, for our customer, Atlas Energy Solutions. This is just the first demonstration that driverless is really here. With that, I'm excited to talk more detail and get into it with Chris. Excellent, Don. I know we have this conversation all the time, but I think it's a great one to have out in public. When I think about that, starting with a very specific unstructured ODD of the three markets that you talked about specifically, you're moving sand, right? In the Permian, now a complete perfect sort of choice and case to start with AV. Could you talk about the idea of why you wanted to start there and then the transition into a more structured environment like highway over time? Sure. One of the, I mean, we started the company as a long-haul trucking provider, right? That was the number one focus. That was the only focus of the company. It's something that I've been pursuing personally for multiple decades. It was really our work with the military that unlocked the potential for other domains. It was not something we, in full transparency, that we actually had intended to go into. Atlas actually came to us. We were not looking at this industry originally. They reached out to us based on the work that we were doing with the military that they had seen for videos and other capabilities that we had posted. We started to think about the challenges that they were facing, the costs associated with their business, which is high even relative to the over-the-road commercial market. It became very interesting and appealing. Additionally, the idea that this is a remote location where you don't encounter a lot of the sort of daily commute traffic or other kind of urban traffic that you would in other domains, that was really appealing. When you're putting your first driverless vehicle on the road, there's a lot of risk. There's potentially more risk, not only from a safety perspective, but also from a credibility perspective for your company. We thought this would be a perfect, both literal and figurative sandbox for us to deploy our solution initially. Also, Atlas really has been a great partner. They've been willing to kind of dig in with us, figure out all the logistics associated with running an actual driverless product. They wanted to own and operate those vehicles, which is something that we've been talking about from a model perspective for a long time. They were willing to take the chance and really pushed for it. That's really the initial reason why we thought it would be a good initial deployment. Now we're scaling up with them. We're learning a ton of lessons on the operational and logistics side, which we think will be applicable to other customers along the way. When we think about the three sort of SAMs that you put up there, could you just give a hypothetical for what we've heard over the course of the day, what on highway, what a software as a service model may be like? What does that look like for something in the Permian? How do we think about how that is structured? I'm sorry, I maybe didn't fully understand the question. Oh, just meaning like how is the monetization for someone like Atlas? You don't have to talk about Atlas specifically, but how you sort of get paid. We have put some ranges of, in terms of the business model, the unit economics. I have to say that the unit economics are very attractive in that industry. That was one of the key reasons that made us look at this as a first and future deployment. As an example, companies like Atlas Energy Solutions already operate their trucks 24/7. If you look at the cost basis to hire enough drivers, they have multiple shifts of drivers in order to operate those trucks. Unlike long haul, where you might put one driver in one truck, and they might make stops along the way, or you have team drivers to keep that truck rolling, these trucks are literally going back and forth day in and day out. It's a 365-day-a-year business that effectively doesn't shut down unless there's some major, major weather event restricting movement of the entire area. That was very attractive because it means that we can go capture and monetize the driver replacement over a higher part of, for a longer period of the day. The average cost in the industry is about $340,000 to move a single truck for a year. That obviously means for a fixed cost of deploying a driverless truck on our side, we can obtain higher revenues. In a world where we have very low deployments and revenues are still very small, we're trying to maximize our revenue potential for every truck we put on the road. Yeah, just we'll finish on sort of that end mark before we move to the transferability to other markets like highway. Atlas becomes an amazing case study, right? Like that it's working now for all of these other markets that you described. Yeah, absolutely. Like I said, it's a sandbox. I've been describing it as there's really three pillars to launching a driverless product. Most people focus on the first one, the first one being technology. Obviously, we all like to talk about the technology and how the technology is different and how the technology is evolving. Really, that's, I think, only one leg of the stool. Then you have the safety case pillar, which is very difficult. At least as far as we're aware, AI is not really solving that problem for us yet, right? There's a lot of experience that comes in proving safety statistically. There's a lot of I's to dot and T's to cross. The third pillar is really the product and building an effective product for a customer, putting it in their hands in a way that they can actually use it effectively and efficiently, seamlessly with all of their dispatching team, all of the folks on the ground to blend it in with their existing driver-in-service, right? Atlas has to deliver. It's a just-in-time business. Every couple of hours, they need to deliver their product to their customers. There's huge penalties and fees if they fail to deliver. The stakes immediately are high. There's no T-ball. You don't get to play in the little leagues before you go to the big leagues. That was something that they really stressed upon us when we embarked on this journey and started working with them. It's like when we accept these trucks, they need to work, and they need to work reliably, and they need to work around the clock. That was kind of a very high bar that they set. The flip side is we get to learn and get experience deploying in that model from the beginning so that when we eventually expand to other customers, other geographies, and other regions, we have that experience under our belt. We know a lot of the gotchas and pitfalls and other things that we may encounter along the way, which we think will assist us to offer a better product to more trucking companies down the road. Is it fair to say in that market right now, for the foreseeable future, you see yourself as really one of the only players, even within people's focus markets, because it is a very specific market? I mean, I won't speak for all of the potential theoretical players and what they're thinking. This is definitely an area that we've focused intensely on over the last, let's say, year and a half. I'm sure that other AV providers are taking notice of the business. I think at the end of the day, competition is healthy. We welcome folks to explore these opportunities if they're interested. Currently, we're not aware of folks who are targeting this space. Let's talk a little bit about that transition from what would be the hurdles from moving from an unstructured environment to a structured environment like on-highway. I think many people would want to know, is that an AI problem, or is that a business model partner implementation issue? We have the benefit of having started in the highway environment, right? Everything we designed and built originally was intended for highway operation. We continue to refine and expand our highway product over time. Despite the fact that we've focused the majority of our safety case and kind of go-to-market on this industrial application over the last 12-plus months, we continue to operate freight for great partners on a day-in and day-out basis for over-the-road application. We continue to collect data. We continue to refine the product. We continue to make progress. We've actually expanded. We've added 3,000 mi to our network over the last six months. We're continuing to push forward there. I would say we've never moved away from that in any sense, in any kind of technical sense. I think the other thing that's worth mentioning is while we do have a defense effort, we have a long-haul trucking program, and we have an off-road and industrial trucking program, they're all the same software. We have one software binary that runs across all three of those. The mission profile may be different, right? What the U.S. Army wants the vehicle to do may be different than what a trucking company may want the vehicle to do. That still might be different than what Atlas needs. The underlying AI technology, the same neural networks, the same AI is running across all three of those platforms. We're not sort of specializing it for one or the other. We don't have internal teams dedicated to these different verticals. We only have one team. We only have one product. We only have one set of data for training for the system. To explicitly answer your question, I don't think it really is a technological hurdle to get through. I think really the nuances in kind of those three pillars where technology is more or less solved, but the safety case for highway operation is something that the company is now turning their focus to. For others who have gone through that process and I think have spoken ad nauseam about the difficulties there, I share that view. It is a long, difficult, tedious road. You have a lot of risk associated with it. You need to make sure that you get it right. We're not going to do anything premature until we know for sure that is absolutely safer than a human. That's going to take some time. Yeah. If we stay on that end market of highway, talking about, like you said, closing that safety case along your timeline towards 2027, one of the questions I've been asking almost all the companies involved in trucking is, what is that end game that you want to be for the first couple of years of operation? We've heard differing views today about hub to hub versus sort of customer endpoints. How do you think about the ideal market in those first couple of years of operation on highway? You're asking what I think the ideal is or what I think will actually happen? Both. Yeah, I think the ideal, pretty much regardless of who you ask, I think the ideal is that you can deliver an end-to-end solution, right? Hubs are not really how the industry works today. I think it's not really how the industry wants to go. We've talked a lot about the hub to hub model. I know others have talked about the hub to hub model. Realistically, I think that is a kind of a temporary or transient solution to the overall problem. We've started working with various customers where we actually go in an end-to-end sense with the technology. I think very quickly you're going to see that that model is the one that's going to ultimately be deployed at scale. I think it's very difficult to make the economics and scaling work for a purely hub to hub model. There will be some application of hubs in some specific areas, but I don't think exclusive hub to hub is really going to be the one that we see scaling into the future. On that AI question for end to end, is it less difficult than we're sort of making it in this discussion? I think some of the stats are like, you know, a lot of the facilities are within, you know, 80% are within a couple of miles of a highway. You're not in dense urban, right? Which is obviously difficult from an AI perspective, but you're on service roads within a couple of miles. Sure. Is it, again, that AI question, is it really more of a focus on closing the safety case versus that AI is not ready to do the couple of miles of service road? Yeah, again, I think this is a transient kind of problem. We talk about what's capable at launch. I mean, we're still in the first inning here. Any launches that have existed are still incredibly small in the grand scheme of things. It's pretty clear now that the technology is improving so rapidly that I think we're going to kind of look back on that question in a couple of years and be like, that was kind of a silly thing to focus on. I'm not so worried about the notion that endpoints are one or two or X miles away from highways, etc. I think that's kind of a problem that most likely most companies are going to easily solve. I think that that will go away. What I will say is very challenging is inside various facilities, like all customer facilities, trucking facilities, they're all very different. They're all very bespoke. Even within the same company, they'll have a lot of different looking facilities in different places. That, I think, is going to be a much longer tail to solve with generalized autonomy all the time, 100% reliably. Even though it's a low-speed environment, they can be very challenging. It is unstructured, right? It can be. It can be in a lot of ways. It's kind of a mix of structured versus unstructured. What you find is that people tend to break the rules a lot in and around parking lots. There's a reason that I think Waymo's deployment didn't make it into parking lots for quite some time after they were doing other things. You can think about customer facilities being very much like parking lots in some ways. That's an area where I think companies are going to find it more difficult to generalize their technology to handle. You know, you can do five, but you can do 10. You can do maybe even 50. Can you do just all of them arbitrarily? That's going to be a little bit more challenging. If I had my shipping colleague here, or if I think about your customers in those first couple of years, what does a typical route look like? Obviously, I know some of the geographies, but distance, is it going to be all about utilization? It doesn't matter if it's a 200-mile route, but you're running it four times a day versus an 800-mile route and you're running it once. How do you think about those, or what are your customers sort of asking you about? There's a wide range of perspectives on that. I think it's a mistake to lump trucking companies all into one mental group, right? They're all very different. Some of them run more routes that are on the shorter side, some of them run longer routes. They value different things. I think it's important to have a conversation and a relationship with every potential user of the autonomous solution to build a model that works for them. I don't think there's a one-size-fits-all answer to this. I'd say in general, longer routes are more economically valuable than shorter routes for autonomy. Drayage. Yeah, I mean, drayage is a completely different thing. I'm talking about long-haul routes that are both short and long. The other thing to consider is that we tend to think of like, oh, yeah, if we could go coast to coast in one shot, wouldn't that be so great? Most of the industry doesn't work that way today. There's a lot of moving parts for us to get to this ideal future where all the trucks are zipping around autonomously at maximum efficiency. From where we are today to that point, there's a lot of kind of milestones that we still need to hit. There's a lot of intermediate distances that are going to need to be tackled. There's going to be, I think, a change in the way that freight models are constructed and trucking companies decide where they are going to have facilities in the future. Of course, all this change takes a lot of time. Nothing happens overnight. I don't think there's a one answer to your question. Certainly, from an autonomy perspective, from a safety case perspective, shorter routes are easier to do. It's easier to prove safety on a shorter route than it is on a longer route because more things can go wrong on a longer route. The nuance is more technical than that, but I think that's a good way to think about it. What you're going to find is that initial deployments are going to tend to be on shorter routes, and then quickly it'll expand beyond that. Don, why don't we move back to it? It's a little bit short-term. It's some of the great things you're doing on execution in the near term. The targeted route dedicated manufacturing line, I think maybe the first truck has now actually rolled off the line. Can you talk about how Upfit works, your bill of materials, and a little bit about that evolution? Because that's, again, to your point, it's all about execution right now. Yeah, I mean, we wanted to find a solution that was viable today. When I say today, we've been working on this for several years, right. Going back several years, there's been a lot of conversations, probably ones that have been had today about the role that OEMs play in this business. I think a lot of it in some ways is theoretical because I live in a world where I need to deliver a product today. I need to find a solution, even if it's not the most ideal solution from some people's perspective. I need to find a solution that is actually viable today if you want to ship a product and you want to start generating revenue. We set out a couple of years back to work directly with tier one suppliers to architect a fully redundant platform that we could then manufacture or hand to a manufacturing partner to make, build, and ultimately sell to customers. Atlas is the first example of that. We brought on Roush Industries. We announced that earlier this summer. They spun up an assembly line specifically for Kodiak Robotics' technology. They are the ones responsible for installing the sensors, our SensorPods, the compute, and kind of putting the finishing touches on the truck. Atlas owns the trucks. They buy the trucks directly from the manufacturer. We're not involved in that transaction at all. The trucks get taken to Roush. As you said, the first trucks have now rolled off the Roush line. That provides us the ability in the short to medium term to scale beyond, let me put it this way, Roush can expand beyond our own capacity for the foreseeable future. We have no more bottlenecks in terms of the trucks that we can deliver to customers. Would it be a better world if everything just came fully baked from the OEMs? Yeah, of course. As far as I'm aware, that isn't available today. At some point, we all believe that it will be available in the future. I'm hopeful and looking forward to that future. We didn't want the future of our business to be in the hands of a timeline that we don't control. Ultimately, that's why we made that decision. We're super happy with the partnership with Roush. I think they're a great manufacturing partner. They provide an amazing product. They support that product. The customer is incredibly happy. Atlas wants to continue to expand and receive more trucks. It's always a necessary evil that I always have to ask at these. Could you talk a little bit about your actual AV kit and the AI approach, the sensors you're using, and whether on the highway you'd use HD maps? The obligatory question. We have a combination of LiDAR, radar, and camera. We've talked about that for a long time. Our current model, our current generation of truck is the same between our industrial application and our highway application. That's valuable because it allows us to collect the same data, to use the same algorithms. Everything's the same across those trucks. The specific chassis used by Atlas is a little bit different than the one you would use for an over-the-road purpose. There are different aerodynamics, and there's different steel construction, et cetera, because it's a very, very harsh environment. It's one of the harshest environments you can imagine operating commercially. There are some minor chassis differences, but for the most part, everything is the same. That's really valuable to us to keep it all the same. I forget what else. Oh, yes, maybe that's HD maps. We don't use HD maps. I've talked a lot about that in the past. I won't belabor the point. I think that modern AI approaches can interpret the world in real time. I think HD maps had a place in time. I don't think they're necessary anymore. I also don't think they're evil, right? They're not doomed if you use them. They weren't a necessity for us in our initial implementation. I do think they would have some challenges in very unstructured environments. That's the one thing about maps in general, they tend to depend on structure because you're mapping something. If you go out into the middle of nowhere where there is nothing to map, then I think that approach is going to struggle. Last question for me that I want to open up to questions from the audience. Can you talk about some of the commercial partnerships that you have in the trucking ecosystem? Just to remind for the 101 out there, some of the main relationships that you're going to rely on for 2027 and onwards. Some of the folks that we work with today? Yeah. Yeah. We've cultivated some great relationships over the many years that we've been in the business. I think our current active partnerships are Werner Enterprises, who's in the audience today. Thank you. J.B. Hunt is another great partner for us. We've worked with some great folks in the past, like C.R. England, the largest refrigerated carrier. Of course, we have Atlas. Bridgestone continues to be a great supporter and partner for Kodiak as well. Excellent. Want to open up to questions from the audience? In the short term, what do you think is going to be a bigger market, be it commercial or military? In the short term, what is going to be a bigger market, commercial or military? That's a great question. The military opportunity is a little bit harder to pinpoint, mainly because of timing. The timing is a little bit uncertain in terms of when contracts are awarded, when programs are initiated. There is a lot of politics involved, Congress and appropriations and budgets and all kinds of things that go into that. That makes it a little bit harder for us to kind of pinpoint what that revenue projections look like. Therefore, from a kind of a public guidance perspective, we haven't guided to anything relative to defense. I remain incredibly bullish on the opportunity that's represented by defense. I think defense is a really important aspect for autonomy. I think autonomy is going to play an incredible role in the future of military and conflict. In the short term, I think commercial is going to provide more of the revenue and a bigger piece of the pie. Other questions from the audience? Don, I always like to focus on, you know, these were my questions. What is your focus? I mean, look, you've had a very, very busy last couple of months. Now that you're public, what do you think is your focus for the next three, six, nine months? I think we finally get to get back to work, which is great. The weekend we could decompress. We listed on Thursday, so it's very fresh. Yesterday was a travel day. This is really the first day I've had since that process, which is great. I think building trust with the community, investment community, partnership community, research community, building trust, right? We need time to actually continue to execute. I think setting realistic expectations and providing reliable guidance is something that I really want to focus on. I've seen a lot of companies go public that have ultra lofty expectations and fail to meet them, and it ends up being a disaster. I think that's a good lesson that we've learned from watching others prior to us. I want to really focus on building a trustworthy relationship with the investment community, continuing to deliver on those milestones, continue to make sure we're expanding our product for our customers, and making sure that that product actually is really valuable for our customers, right? Because without the ability to deploy the product, then we don't really have anything. Focusing on the customer is always very important. Setting realistic milestones for the company and being honest and transparent about where we are and where we're going, I think is going to be important for us in the next several quarters to years. OK. Excellent. With that, a round of applause for Kodiak and Don Burnette for going public. Here we go. Excellent.