Thank you, everyone. I think we're going to keep going. Good morning. My name is Adam Samuelson. I'm Senior Vice President on the Aerospace and Defense Equity Research team here at Jefferies. Welcome, everyone, to our annual Jefferies Global Industrials Conference 2026 in New York City. I would say we are lucky enough to have Ben Wolff, who's the CEO of Palladyne AI, joining us. Ben served as the President and CEO of Palladyne since February 2024, has served as a member of the board since September 2021. Prior to joining, he was the CEO, President, Chairman at Pendrell Corporation and co-founded Clearwire Corporation, where he was the President and CEO. At Palladyne, he is leading the company strategy across decentralized, embodied collaborative autonomy, defense robotics, AI-enabled UAV software, low-cost autonomous systems, and U.S. based defense manufacturing. Hope I didn't miss anything.
You got it all.
in that list. We're going to just jump right into questions, and there's time later at the end, happy to take questions from the audience if there is any. So maybe first, Ben, thank you for joining us. Just to start, for a high level for investors who are newer to the story, just frame Palladyne AI today. How do we think about the mix between autonomy software, defense hardware, precision manufacturing, and the commercial IQ pilot businesses?
First and foremost, we are an embodied AI company. That is really central to everything that we do. And embodied AI means we make artificial intelligence software that sits on machines to make machines smarter. It's not something that sits in the cloud. It sits on the machine and enables the machine to be able to observe, reason, and act much in a similar way to the way we humans do. So that's central to what we do. Beyond that, we have weapons systems that are designed from the ground up to be able to leverage our AI. And we also manufacture components, both avionics and precision components, that are designed to be built into weapons systems that leverage our AI. So we are both a defense prime.
We are also a subcontractor to other defense primes, and we are happy to sell our software and our products both directly to the government and to other primes.
Got it. Let's think about that in kind of where you sit today. On the Q2 call, you said you were more bullish on the company's prospects than any point since you've come back into the CEO role. Is there anything specific in terms of customer traction, backlog, DOW kind of military exercises and program visibility? What would you attribute that to?
I came back to the company 2 and a half years ago, and we did a pivot then, and we decided to focus primarily on our AI software, and we started planting a lot of seeds. I don't know if you're a gardener, but I like the analogy. We planted a lot of seeds back then. In year one, we had little shoots that came up, and we were encouraged, but there wasn't a lot to be able to point to publicly and say there was reason for a lot of enthusiasm. We're now going on, this will be coming up on my third year with the company, back at the company. I can say now we're getting a serious crop to harvest. We're starting to do some of that harvesting.
Across every one of the initiatives in the company, whether it is our autonomous swarming software for drones, whether it is our industrial robotics AI software to enhance manufacturing, whether it is the drones that we've developed internally, or our avionics and precision components business, we are seeing significant interest across the board, and we're generating meaningful revenues compared to where we were historically. I feel like it's harvest time, and it's across the company, that's why I'm as enthusiastic as I am. Most importantly, back to the AI theme, we have an entirely different type of AI architecture than anybody else has tried to deploy. That is a leap of faith that we had to take, that our customers had to take, and that our investors had to take, and we're seeing it pay off.
So much so now that we have more calls coming in from potential customers than we are making going out. To me, that's a kind of a tide change, and I like that.
Okay. That's helpful. So maybe just kind of on some of that momentum, in the Q2 , you had revenues up 63% sequentially to $5.8 million, backlog about $25 million. You kind of reiterated your full year revenue guidance, $24 million to $27 million. There is an implied second half ramp in there. Just how do we think about the visibility that you have to that? What's contracted backlog? What are awards that you still need? Help us think about how we build from a very solid first half.
I have to say this up front. My lawyers say that I can't reiterate guidance because of where we are in the quarter, so I won't get to the specifics of the guidance. What I will say is when you take a look at the $13 million of new contracts in just the Q2 , where that put us from a backlog perspective, the fact that we don't talk about backlog, and we don't include numbers in backlog unless it is contractually committed. So it is a binding legal commitment from our customer. You look at that kind of ramp, and we are just in the first inning.
I think when I look at where we are from kind of the progression of that business and those shoots coming up that I was talking about ready for harvest, I think it's all trending in exactly the right direction.
Got it. Maybe that is a good segue. A key investor question is thinking about how you transition from some of the demonstration and trial works that you have had underway to actual production revenue. How do we think about that transition from trial to production, and which parts of the company are actually furthest along in that?
Most of the trials or demonstrations that we have been doing for the last year or so have been on the AI side. That is the high margin software business. As we scale that does not require more CapEx. It does not require a lot more OpEx. Once we created the software, we start stamping it out, sending out copies of it. That is what our primary focus has been. We have had a number of successful exercises with the U.S. Army. We have had a number of successful demonstrations with the Air Force. As I said, that is now yielding inbound calls to us. Word is spreading throughout the Department of War about our capabilities, and we are getting inbound calls from folks we have never talked to or heard of before. That is extremely encouraging.
We are getting glowing reviews on the way the product is performing, and most importantly, we are being told, of course, this is not statistically significant in any way, but we are being told that nobody else's software does anything close to what ours does.
Maybe digging into that a little bit more, you have had some notable successes with the Army, including IV exercises, IVAS, PC, C6. Just what did the company learn about its technology in real-world operating conditions? How does that feedback translate into the future program requirements and business opportunities?
One of the big things that we didn't know before we started these exercises was whether the system, the software, would be simple to use. We believed it was, but when you have a lot of geeky engineers putting together software, you can go one of two ways. You can have a phone that looks like an early Samsung smartphone, or you can have a phone that looks like an early iPhone. Obviously, there was a significant difference in usability of those two things. You look at the features and functionality of the early Samsung phones and nobody used 75% of those features and functionality because they weren't easily accessible. They didn't understand how to use it.
We started deploying it so that we could figure out, can a low-level E-1 soldier get trained on this software in a matter of minutes and start flying drones? Not just flying a single drone, but flying a whole swarm of drones with just a single soldier in control, which has never been done before. We put it in the hands of soldiers, and we saw over and over again that folks that had no meaningful drone experience could use our system, learn in a matter of a half hour how to use it proficiently, and fly drones. That was the big test for us.
We knew that it was capable of flying swarms, and to make those swarms, each drone in the swarm fully autonomous, able to make decisions in flight in real time and respond to what it was seeing and sensing on the ground. But it took taking an E-1 soldier, 18-year-old E-1 soldier, actually putting his or her hands on it to be able to demonstrate that it was usable in the field under pressure.
I guess tied to that is this question of vendor neutrality, right? A lot of the drone manufacturers, in their own ways, have their own software platforms or their own preferred providers. Where you sit, where you don't have your own, at least not right now, major kind of drone platform internally, how does that opportunity translate into the DOW requirements for multiple different vendors and collaboration across different suppliers?
We designed our system to be used across any type of drones that are going to fly. It is not specific to a certain platform. Obviously, some companies that are out there talking about autonomous capabilities on drones, they are focused on enabling their own homemade drones to be able to become autonomous. We designed the system architecturally from the ground up to be able to enable a heterogeneous swarm of drones, meaning drones from any and all manufacturers that have our software on it would be able to collaborate autonomously in the air and affect the mission outcome. That is central to our thesis.
What we saw in the exercises that we were doing, where we took drones from four different manufacturers, put our software on it, now a single soldier could control the whole mission for all four of those drones, that is exactly what we wanted to prove. Of course, that is very appealing to the military. They do not want to get vendor lock, as they say, on a specific hardware platform, particularly in this world where drones are evolving so quickly. What they want to use today might not be what they want to use tomorrow from a hardware platform perspective. To have a software platform that could be consistent across all of those different platforms, you have a soldier learn it once, and that is it.
In the early days of the PC, we are kind of like Microsoft. We will put our software on anybody's PC. Sometimes you make your own product to really show what your software can do, as Microsoft did from time to time. But ultimately, they relied on the big ecosystem of PC manufacturers that they could then put their software on.
Got it. Okay. Maybe we will stay on the drone topic, but from a different perspective. A couple of months ago, company announced a partnership with IAI to Americanize the Harop, Harpy, and Mini Harpy for the DOW with Palladyne. Can you just give us an update on how that has progressed since the announcement in early June? How you think about what you have learned since then about the size, urgency, and shape of U.S. demand for loitering munitions.
IAI, Israel Aerospace Industries, is the largest defense contractor in Israel. They are entirely government-owned today. They selected us to be their partner in the U.S. to bring to the U.S. three of their loitering munitions platforms. These loitering munition platforms are large platforms. When I say loitering munition, just so that everybody understands, we are talking about something that is a hybrid between a UAV and a missile. It looks like a big UAV, 10-foot wingspan, but it has a warhead on it. It has the ability to loiter for up to 9 hours, go up to 1,000 kilometers in range. Its real claim to fame is it is specifically designed to be able to destroy enemy air defenses, radars specifically. It has a sensor on it.
The Harpy and the Mini Harpy have sensors on them that enable them to detect brief or intermittent radar signals from the ground. As soon as that radar pops up and starts emitting, trying to detect our incoming jets, that loiter munition can detect that radar signature and destroy it. We do not have anything in the U.S. arsenal today that is able to complete that mission. The way we solve that problem today is when we know we want to fly jets into a theater, we have other jets with pilots that are specifically focused on launching missiles from those jets to destroy the enemy air defenses.
The more we have enemy air defenses that are mobile in nature, they are hidden in a cave somewhere or in a warehouse, and they pull them out, pop up the radar dish, shoot at our jets, and put them back in their hole. That gets really tough to defend against. This is the only solution that I am aware of that can solve that problem. We do not have anything else in our arsenal that can do that. That is why we were attracted to it. It becomes particularly attractive when we can have a swarm of those types of munitions in the air, all collaborating, sharing sensor data, sharing targeting information, sharing information about what the enemy is doing on the ground. That is why we got very interested in that.
Our objective is to take those Israeli-made machines, Americanize them, modify them for whatever the Department of War's requirements are, and produce them at scale here with our software on them. What have we learned since we announced it? It has only been not even 90 days. What we have learned is that there is, in fact, nothing else in the U.S. arsenal that can do that job. This is by far the most cost-effective solution. To be clear, these are not cheap drones. These are not Group 1 $20,000 drones. These are exquisite, expensive systems, but highly capable and able to complete that mission at a small fraction of the cost of scrambling jets and shooting Hellfire missiles or whatever other ordnance we might shoot.
We've learned that people in the military are aware of these systems, that there is real interest in having that capability here, but it's early days.
If there isn't that capability set yet in the U.S. arsenal, is there an existing program of record or funding vehicle that you could see that could purchase these systems if they are made with the right U.S. supply chain, or do they need a new program to be created?
There is funding that would enable relatively small purchases to get started. These are contracts that will be hundreds of millions or billions of dollars. At that size, if you wind up talking about them being purchased at scale, we'll need a new program of record.
Okay. That's probably one of the bigger gating factors.
I think for us, the IAI relationship is very much binary. It is not going to be a little bit here and there. Just because of the size and scale of the program, what we will need to do to stand up manufacturing capability, et cetera. We have made it very clear to the government that we will stand up manufacturing capability as soon as they are ready to buy, but we are not going to spend a lot of our investors' funds on something that is a build it and they will come type of mentality.
Okay. All right. That is very helpful. Maybe staying on some of the military applications. How do we think about Palladyne's exposure to major missile and missile defense programs? There has been some very large multi-year framework agreements on programs like that and PAC-3 with the primes. You do supply some components and different subsystems into that arena. How do we think about the company's exposure as some of those production rates are set to move very meaningfully higher?
We are a beneficiary of some of that. Our components business participates in some of those programs, and we are winning more all the time. You can expect to see us continue to grow that part of the business. That is an important part. Everything that we do with the major defense primes, it is kind of an entrée or a gateway into being able to do more with them, but we start with precision components. That includes a lot of the most interesting work that I think we do is in the propulsion systems for missiles.
Got it. Any way to quantify the size of that business today, or content on some of the major programs that you have seen some of these framework agreements and what the capacity of the company is today? Clearly there might be some capital might have to go against growing that production set going forward.
Today, we are operating at only about 30% of our total capacity, and that means, without hiring a lot more people and without putting any more CapEx in the ground, we could triple our production, roughly, of our components business and not have any more capital required. That's a good place to be forward-looking. It's not the greatest place to be today. We've got an opportunity to really crank up that part of the business, and that is what we are focused on doing.
Got it. Any way to maybe help people out triangulate, kind of think about the major programs, the ones you're most exposed on, or too far fetched?
We are restricted in what we can talk about. But, I would say that when we are allowed to publicly announce things, even if it is just generically in terms of size and quantity, we will do that.
Okay. That's helpful. Maybe switching gears to the commercial and industrial side. Maybe we can start with some of the news that was announced I guess it was yesterday.
Yesterday.
There was an announcement with FANUC on the industrial automation side. Can you provide an overview of what that relationship looks like and how the company's positioned there?
Yeah. FANUC is obviously one of the largest manufacturers of industrial robots, and has, I think, the largest market share in the U.S. Very embedded in a lot of the big auto companies and really across the board on industrials. Our software can get loaded onto a FANUC robot. It does two key things. Number one is it allows that robot to be trained to perform a new task in a fraction of the time that it has historically taken to write millions of lines of code to enable a new application or task. We are able to take a person that's a line worker, for example, not a computer programmer, a line worker, and have them, in a matter of minutes, show that robot how to perform a new task without having to write any code at all.
We call it a low-code or no-code solution. That means that the robot can get stood up and put into revenue-generating business faster than it's ever been done before. We can shorten months to minutes. That's a big deal. That enhances the ROI for the ultimate customer. The second thing that the software does is it enables the robot to be able to deal with variation or variability. Let me give you one example. One of the tasks that we've got a robot doing today is painting or preparing a bicycle frame for painting. It requires a primer coat, basically. In the production facility, you've got these bicycle frames that are swinging overhead on an assembly line. As that frame goes down the assembly line, it's always rotating a little bit, and then you're trying to apply paint to it.
That's the kind of application where that variability of what we call the pose of the object being manipulated always changes, and it's impossible for traditional automation to be able to paint that effectively. We're able to enable a robot to do that. So it's a very real use case. It shows the variation in both the environment and the object, and we're now able to automate that task. That's the kind of thing that we're doing with FANUC. Basically, we're taking jobs or tasks that historically have been so complex only humans can do them, and now we're enabling them to be automated using a FANUC robot. That creates, obviously, a lot of interest by FANUC because it will allow them to sell more robots. It opens up markets they wouldn't have previously had the ability to attack.
Got it. As you think about that, obviously the announcement was made yesterday, that is going to take a while for that to season, but did you envision FANUC becoming your principal channel partner and effectively distribution into that industrial automation space to-
We have hoped for them to be a channel partner for us for the last year or so. We have been working with them closely for about a year. They finally had enough confidence and positive customer feedback that they said, "All right. Let us make this official. Let us make it public." They are now promoting us, actively promoting us. We are doing joint sales motions with them. It is a great thing. Whether they wind up being our primary channel, I think it is too early to say primary versus. Just as we talked about on drones, we are robot agnostic, so we work with other vendors also. Our software is capable of working on virtually anybody's robot.
We will continue to push the pedal on all of them, but certainly having FANUC, who is without question a great leader in this country and outside of this country on industrial automation, as a partner will be huge for us.
Just conceptually at least, this is a software license business from how you would go to market and margin profile more like a software subscription.
You bet. Think about it, again, back to the early days of Microsoft. Microsoft winds up with a partnership with Dell, and yeah, they have their operating system on the robot or on the computer. That is what it gets shipped with a FANUC standard, just like the early days with MS-DOS on a computer. But then if Microsoft had a customer that wanted to buy, or Dell had a customer that wanted to buy Word and Excel and the whole Office suite, they added that on back then. That was an add-on. It was not just a gimme. Think of us as being a value-added software provider along that same line. Some robots will be shipped with our software on it and some will not.
Okay. Maybe beyond the FANUC relationship, can you talk to me or can you give us some color on where else you are on the commercial industrial side on go-to-market and applications that are closer to scaling into more meaningful revenue opportunities?
That Palladyne IQ industrial robotics product is our primary focus on the commercial side. On the drone side, we do a little bit. There are certainly opportunities with public safety and border patrol and things like that. But when you think about our ability to materially enhance the defense industrial base on the production side, IQ is our bet, and we are getting great traction there.
Okay. Turning to maybe profitability and cash generation, you have kind of reiterated your full year cash burn guidance. You do expect the cash burn to decline in the second half of the year. As we move beyond the second half, at least conceptually, just how do we think about the levers in the business to move margins and cash generation to higher levels? Is it really just revenue scale, which brings with it manufacturing utilization? Is it the mix with software? Is there room on the OpEx side, working capital? There is a lot of different levers inside the company. How do we think about which one moves the needle the most?
I think additional revenue with contribution margin is what moves the needle the most. We can scale the business dramatically, whether it's taking up additional capacity on the manufacturing side or selling software. In both of those key areas, it doesn't take us hiring more people or putting more CapEx in the ground. We have the opportunity now to really hit the accelerator, scale the net revenues coming in, and every dollar of gross profit is going down to the bottom line without additional cost.
Right. We talked about the AI relationship would require—
That one would require some expansion of manufacturing capacity.
Yeah.
But again, I view that as success-based only. We're not going to put capital in the ground on that one unless we have the government saying, "Here's a very large contract for you.
Got it. On the precision components for different type of missiles and propulsion components, talked about-
We've got capacity
up to 3x. The software side, similarly, that's-
Software side, it's infinite.
Yeah. OPEX even wouldn't scale, would they?
That's correct. Really, you're talking about sales and distribution, and that's it.
Yeah.
Our R&D is our R&D. We're going to continue to spend the same amount that we historically have on R&D to keep us at the forefront of the state of the art of the technology. As we continue to increase revenues, that goes to offset and buy down that R&D cost.
Got it. I want to make sure if there's any questions in the audience. Just give me a microphone if you just wait one second . T his is for the webcast. Sorry.
Oh, sure.
It's coming.
It's coming right there.
Hi, how are you?
Good.
Nice company. Quick question. Do you take an ontological approach with your AI and the software?
We do not.
You do not. How being, natural question would be, you are looking to be hardware-agnostic, cross-platforms, I get it. I am in the space. My question is, how do you learn from the different nuances of the different products versus taking an ontological approach where you could glean a lot of that intelligence and then apply it across other products?
Our architecture, without getting too much into the specifics, is focused on prediction and game theory, and that basic architecture works regardless of the type of machine, the sensors, or whatever we have got on the actual machine. We have to have sufficient compute. It is very much focused on algorithms that are intended to figure out what is likely to happen next using a game theory approach.
Okay. Second part of a question, a little bit different. Some of my companies that I am invested in or shareholders in that are involved in the defense tech space, there is a lot of people shifting jobs, getting fired in government, Army, Air Force. Have you run into obstacles with the people who say yes for the next contract? Are you trying to like, does it step people, I am hearing this. It steps guys back, and then they have to go back and redevelop the whole cycle again. Are you running into that at all?
Yeah. The turnover at the Department of War creates real challenges. You might spend six months cultivating a relationship, getting people up to speed, understanding what the opportunity set looks like, making sure you're filling a real need, and then they're gone. There's no question, then you're back to square one again. Sometimes you're not even at square one because there may not be somebody in that same seat that is even empowered to learn or make decisions. There's no question that the churn and turnover has been a challenge, especially on the contracting side.
Thanks.
Okay. Thank you. Part of this, across many of our presentations here, we're going to try to ask a series of similar questions to all the different companies that we cover. Maybe the first, just high level, how are you thinking about your own 2027 revenue growth potential as you sit here today?
As I mentioned before, I think we're in the first inning. It's been a long inning, but it's the first one. I see nowhere to go but up in terms of what 2027's going to look like compared to 2026. We are firing on all cylinders right now. I'm not going to give any forecast or put any numbers behind it, but there's a reason why I said on our last call that I'm feeling more optimistic than I have at any time since I returned to the company. I think that we are in the right place at the right time across our entire portfolio of products and services.
Okay. Second question is there any notable bottlenecks in your current supply chain from a parts, whether those are electrical, electronics, mechanical, anything that you'd be calling out right now?
I'll tell you, the biggest thing that I'd call out is we're having a hard time getting enough computers for our engineers, and it's because of the chip shortage. The cost of compute and availability of compute has become so constrained. There's a follow-on effect of that. If and when we wind up seeing significant volumes of drone purchases, we've got to make sure we've got sufficient sources of the compute. We're in a little better position than most because we're making our own avionics suite. We have no concerns right now with what we're seeing, but we're having to buy earlier to make sure we're covered. Also again, the price volatility. I shouldn't say volatility. It just keeps going up and up and up. There's no down.
The volatility both ways might be a benefit at some point. Along those lines, any suppliers consistently underperforming?
We fortunately are not in a position today of having underperforming suppliers. Realize that we're just getting started in the volume business. Right now, we are feeling very confident. We hired a true world-class person to lead our supply chain efforts, and I think we've got that well under control. You can always, in this environment, you can always have a curveball thrown at you.
Got it. I think maybe you already answered this, but where are you experiencing the most significant price inflation?
It's on the chips, on the compute.
What are your current thoughts about the labor market?
We're fortunate in our business because we have cool, sexy stuff to work on. We have very little challenge hiring engineers that we need. That's really the biggest concern in our industry, is availability of engineers with the kind of background and experience that we need. Labor market generally, though, is tight everywhere and there are fewer and fewer people wanting to go into blue-collar jobs. Young people don't come out of high school saying, "I want to take a blue-collar job." That is what creates the opportunity for us on the manufacturing floor with our Palladyne IQ product. I don't want to say that we rejoice in other people's misery, but we are bringing a solution to a big problem to market.
What do you worry most about in 2027?
Unpredictability. It ties a little bit to the question that we got from the audience. Continuing change and upheaval, whether it's geopolitically, whether it's the Department of War, that always interrupts things. It interrupts contracting, it interrupts what the priorities are. We're in a business that we're trying to play the long game here, and if things are changing every 90 days or 60 days, it can delay things. It can cause you to have to change priorities to respond to what the customer's looking for. It's just the uncertainty.
Okay. Finally, last one. Any thoughts on the defense budget?
In the words of our fearless leader, it's going to be huge. Huge. It doesn't really matter, the specifics, how many hundreds of millions of dollars one way or another, I'm not terribly worried about. It is going to be the biggest defense budget we've ever seen. It will continue to be, it's going to continue to grow, and most importantly for us, they're prioritizing spending in the areas that we're focused on.
Okay. Maybe time for final question. This is in the survey. This is generally to wrap up. If the company continues to execute on the pipeline in front of it, what does Palladyne AI look like three to five years from now? What do you think investors are most under-appreciating about the story today?
There's a lot of loaded questions in there. If five years from now we are not a large multi-billion dollar company with the best margins in the industry, I will be disappointed.
Okay. That's-
What are the investors missing?
I think they're missing that.
Well, that's just a self-interested CEO bragging about how successful I think we'll be. Beyond that, we have a true differentiated AI architecture, and that, as some recent people described out of the military, is a national treasure. My goal is to prove that.
Great. Well, I think that's a great note to end it on. Ben, I want to thank you for joining.
Great. Thank you for having me.