AEye, Inc.

Ladies and gentlemen, welcome to AIQ2026 conference call. Call lines have been placed on mute to prevent any background noise. After the speaker's remarks, there will be a question and answer session. If you would like to ask a question during this time, just press star followed by the number one on your telephone keypad. And if you would like to withdraw your question, press star one again. And please note that this call is being recorded. I would now like to turn the call over to Keaton Olson. You may begin.

Good afternoon, and thank you for joining AI's second quarter 2026 earnings call. I'm Keaton Olson, investor relations manager for AI, and with me today are Matt Fisch, chief executive officer, and Conor Tierney, chief financial officer. Earlier today, AI announced its financial results for the second quarter ended June 30th, 2026. A copy of the press release is available in the investor relations section of the company's website. Before we begin, today's discussion may include forward-looking statements as defined in the securities laws and regulations of the United States with reference to future events, operating results, or performance, and are based on our current expectations and assumptions. Any forward-looking statements are subject to inherent risks, uncertainties, and changes in circumstances. Our actual results may differ materially from those contemplated by these forward-looking statements. You can find more information about the risks, uncertainties, and other factors in the reports AI files from time to time with the Securities and Exchange Commission, including in our most recent periodic report. The statements to be made are as of today only, and AI does not intend to update any forward-looking statements regardless of any new information, future developments, or otherwise, except as may be required by law. In addition, we will be discussing non-GAAP financial measures on this call, which we believe are relevant in assessing the financial performance of the business. These measures are presented as supplemental information only and should not be considered a substitute for financial information presented in accordance with GAAP. You can find reconciliations of these metrics to the most directly comparable GAAP measures within the press release. With that, I'll hand the call over to Matt.

Thank you, Keaton, and thank you all for joining us. I'm excited to report that our overall momentum continued in the second quarter, including the ongoing development of our sales pipeline, having grown to our highest level of engagement yet. And most importantly, I'm pleased to announce that we have been selected as the preferred LIDAR vendor for groundbreaking sports analytics provider. As a company, we're hitting our stride, and we extended AI's reach into an entirely new market category. Q2 revenue grew approximately nine times year over year and roughly 100% quarter over quarter, marking our fourth consecutive quarter of growth, further evidence of the accelerating commercial interest in our technology. Beyond revenue, our sales funnel continues to be the best barometer of our progress. Proof-of-concept programs from revenue-generating customers grew to 25% from 21 since our last earnings call, and both engagements and quote activity increased approximately 25 and 40% quarter over quarter, respectively. Commercial momentum is accelerating across the board. New accounts continue to enter and move through the funnel, and existing customers are scaling up the scope of their engagement. AI sits at the center of the physical AI opportunity. a market in Barclays sizes at up to $1 trillion by 2035. LIDAR is what gives machines sight, and our software-defined architecture positions Apollo and Stratos as core perception platforms across automotive, trucking, aerospace and defense, rail, infrastructure, ITS, and now sports analytics. Defense continues to be our most active vertical, with engagements doubling quarter over quarter. Our lead defense customer placed its third consecutive paid order this quarter, and repeat business is emerging as Apollo is evaluated across UAV, UGV, and counter UAS applications. Additionally, our partnership with Syntec, a leading international defense systems company, continues to actively promote and ship Apollo to its customers, expanding our addressable pipeline into global markets, where Apollo's configurability and long-range performance make it purpose-built for mission-critical and unmanned systems. Increasingly, our customers are finding us rather than the reverse. and often for novel applications that continue to expand our TAM. What makes this possible is our software-defined architecture. Because Apollo's behavior is defined in software rather than fixed in hardware, we can quickly reconfigure it to meet a new application without redesigning the sensor. Customers come to us with a problem, and we can adapt the same platform to solve it. Sports Analytics this quarter is a great example, a new use case addressed with our same underlying product. For our customers, that flexibility is what turns LIDAR-based perception into a practical tool for improving safety, lowering total cost of ownership, and opening new revenue streams. And for AI, it means each new application expands our addressable market without the need to make physical changes to our hardware. Alive3D recently selected Apollo as its LiDAR solution for next-generation sports analytics. Apollo's software-defined technology will enable Alive3D to deliver 3D spatial sports visualization, precise measurement, and advanced data analytics. We are able to attune Apollo to meet the needs of our customers in this case, Elite Sports. This is another clear example of where our software-defined architecture allows us to tailor our solution to deliver for our customers. And Apollo was recently validated on NVIDIA Drive AGX Thor, NVIDIA's next-generation automotive and physical AI compute platform. deepening a relationship that spans our participation in the NVIDIA Halos AI Systems Inspection Lab and our existing validation on NVIDIA Drive ADX Oren. With sensor-to-compute interoperability confirmed against NVIDIA Drive OS, placing AI as a sensor partner in the NVIDIA Drive Hyperion ecosystem, OEMs and Tier 1 suppliers building on the NVIDIA Drive platform can source Apollo as a pre-qualified sensor, reducing integration risk and shortening the path from system design to deployment. Automotive and trucking OEMs continue to treat long-range LIDAR as essential, not optional, for highway ADAS and autonomy. We remain active in multiple OEM Level 3 and Level 4 evaluation projects. In parallel, we signed an MOU to explore combining Apollo's long-range 3D object detection with Movilwheel's acoustic road surface friction sensing. This partnership seeks to provide real-time predictive friction coefficients of road surfaces to improve ADAPT and autonomous driving in adverse weather. Evaluations are underway across select geographies and we are already having discussions with automotive OEMs about potential applications. In other markets, our Optus platform continues to move into deployment. Our live smart intersection in the Bay Area remains operational, as do our multiple Optus installations in and around Detroit. In APAC, Apollo received the Smart Sensing Technology Innovation Award at the EAC2026 Jiao Awards in Shanghai, validating its role in intelligent sensing for ADAPT, autonomous driving, and physical AI application. Additionally, we are now shipping units to an ITS customer we met during our Q1 Roadshow in Korea, and our China partnership with ATI remains actively quoting today. Our manufacturing footprint and capital light model underpin all of this. As an American company with a globally diversified supply chain, we're positioned to navigate geopolitical risk and shifting trade policy better than peers. Through LightOn, we have a dedicated production line capable of up to 60,000 Apollo units annually, derived from off-the-shelf telecom components for mass manufacturability, at competitive cost. We have begun ramping our output to match our forecast of increased customer demand in the second half of the year. Finally, for those customers that require end-to-end perception solutions, our Optus system is underpinned by a partner-led model spanning NVIDIA, FlashEye, Blueband, Black Sesame, Vuron, and now MuvaWheel. This lets us deliver those solutions across market segments without absorbing the cost and balance sheet impact of building every capability in-house. A structural advantage peers with internally developed software stacks can easily scale. With that, I'll turn the call over to Conor to walk through our financial results and the conversion metrics behind this momentum.

Thank you, Matt. Our active customer base reached a new level of diversification this quarter, and repeat business is now a pattern rather than an exception, which is the strongest indicator we have of product market fit and a direct validation of the performance advantages of our software-defined architecture. Before I turn to the numbers, it is worth framing what we believe is happening in our market, because it explains where our revenue is now coming from. The first wave of LiDAR adoption competed largely on cost and packaging. A second wave is now forming, and it has been decided on something different. Performance and sophistication are the gates to winning the deal. The customers driving this wave and creating novel demand in the physical AI space were absent from the first one because their use cases were simply too demanding for first-generation sensors. Protection at extreme range, maximum ruggedness, and a centimeter level capture across the entire playing field. Apollo's ultra long range performance and software configurability paired with our highly scalable partnership and production models put us in a uniquely strong position to capitalize on precisely this class of physical AI customer. We now have commercial proof that this is more than a thesis. During the quarter, we secured a commercial program with a live 3D and a new vertical that was not contemplated a year ago. One uncapability rather than price. Behind this program, a growing number of engagements are moving out of proof of concept and into commercial closure. Turning to the financials, second quarter revenue was $202,000, roughly double the $101,000 we reported in the first quarter. on approximately nine times the $22,000 in the second quarter of 2025. First half revenue of $303,000 already exceeds our full year 2025 revenue of $233,000. Two components drove the quarter. The first is product revenue with an increasing share coming from repeat orders rather than first-time evaluations. Repeat orders from our lead defense customer continue to grow in size quarter over quarter and we started shipping units to Alive3D. The second component, and new this quarter, is $30,000 of contract development revenue from customer-funded engineering work. That line matters more than its size suggests. It is a second distinct source of revenue that did not exist for us six months ago, and we expect it to become a more regular contributor in subsequent quarters as programs advance into their engineering phases. GAAP operating expenses were $10.6 million versus $8.9 million in the first quarter. More than half of that increase was non-cash stock-based compensation, primarily from performance-based equity awards. The balance was non-recurring engineering, tooling, and test costs as we ramp production capacity ahead of anticipated demand. Non-GAAP operating expenses, which exclude stock-based compensation were $8.2 million compared to $7.4 million. I would characterize this step-up as investment ahead of volume rather than an increase in our underlying run rate cost structure. We reported a GAAP net loss of $10 million or 22 cents per share compared to $8.3 million or 18 cents per share in the first quarter. On a non-GAAP basis, our net loss with $7.6 million or 17 cents per share versus $6.7 million or 15 cents per share in the first quarter. Second quarter cash consumption with $7.5 million compared to $9.2 million in the first quarter, reflecting a decrease from one-time payroll costs in the first quarter, partially offset by payments for professional fees, non-recurring engineering costs, and inventory purchases in the current quarter. We ended the quarter with cash, cash equivalents and marketable securities of approximately $71.5 million compared to $77.2 million at the end of the first quarter of 2026. We're reaffirming our 2026 full-year cash use outlook of $30 to $35 million, inclusive of approximately $5 million in working capital. I would note that we expect second half consumption to run higher than the first half as we start to ramp the manufacturing build. This outlook reflects continued investment in commercial execution. The sales coverage, partner support, and deployment infrastructure required to convert our pipeline into a meaningful and durable revenue ramp. Our capital structure also remains simplified and strong. with AI virtually debt-free. That matters directly to the OEMs and industrial customers we're targeting, where multi-year program confidence is a prerequisite for selection. The architectural advantage of our sensors Matt described earlier compounds at the financial level. The same software-defined platform that lets us tune Apollo and Stratus to customer-specific requirements is what lets us enter new markets and solve new customer problems. from defense to automotive to sports analytics without rebuilding our stack each time, when Stratus extends performance into new tiers without a proportional increase in development costs. Piers with fixed function sensors and internally owned software stacks can't replicate that flexibility without absorbing significant development and integration expense. Our expectation for 2026 remains unchanged, As engagements convert into program commitments, we're building the foundation for a meaningful revenue reflection. And we believe we can reach it without outspending competitors. The difference this quarter is that conversion is no longer theoretical. I'll now turn the call back to Matt for closing remarks.

Thank you, Conor. As we enter the back half of 2026, our focus doesn't change. to convert engagements into deployments and continue to turn the physical AI tailwinds we're seeing into a durable revenue ramp. We're entering that period with new proof points. Apollo validated our NVIDIA's drive AGX4 and a live 3D selecting Apollo for elite sports analytics, extending our footprint into transportation and stadiums, all from one software-defined architecture. Our technology continues to differentiate. Our balance sheet provides the stability to execute, and the partnerships we've built from NVIDIA and LightOn to a live 3D lay the foundation for scale across multiple sectors. Operator, we're now ready to open the line for questions.

We will now begin the question and answer session. If you would like to ask a question at this time, just press star followed by the number one on your telephone keypad. and if you would like to withdraw your question press star one again. Thank you. And your first question comes from the line of Richard Shannon with Greg Halem. Richard, please go ahead.

Well, hi, Matt and Conor. Thanks for letting me ask a few questions and congrats on a really nice quarter, some good growth here. Got a few questions for you and the first comment I'm going to touch on here was a very interesting one about ramping output here in the second half of the year. obviously everyone, since you didn't give any sort of revenue guidance, everyone would like to get a sense of what kind of ramp we're talking about, any way to quantify or at least characterize that would be helpful here and maybe any details on any end markets or specific customers you have in mind here for a lot of this ramp.

Hey, Richard. Hey, I'll start it off here and we'll let Conor jump in. Hope you're having a great summer and thanks for joining us today. Look, we're take a look at the company a year ago. We're running around the halls finishing product and working on product, and today we're fully loaded with 25 different customer programs. These are all customers who bought sensors and other things from us, and such a different place. One of the main things we see here is that we talked about pipeline a lot, so here we are at this record pipeline level, but it takes time. and I'll say measured in months to march things through the pipeline and get things working the way the customer likes. And Alive3D is the first one that we were happy to talk about to come out of that pipeline. And the pipeline's pretty deep. We've got stuff coming in, as we mentioned earlier. So we expect that to continue through the back half of the year. And the way I'll sum it up before I hand it off to Conor here is we're just focused maniacally on quarter-on-quarter growth. with the revenue output. We expect that to continue. You've seen that we've managed to run that playbook the last four quarters. We've doubled quarter-on-quarter consistently, and we aim to continue down that path here in the back half of the year, and that's what we believe we can achieve. And, Conor, anything you'd like to add on that?

Yeah, maybe I'd just say, look, we've said in the past that we would be very cautious about ramping production and we would make sure that that's gated to customer demand. So I think that's an indication that we're confident in the pipeline and we're confident in the opportunities here that are kind of moving further down the funnel. And obviously we have 25 customers in play here. We announced one commercial win, but we think we're on the one yard line here with a few other customer announcements. and, you know, we expect to probably be getting some news out here in the next few months or so. And so I think that's what's really driving that ramp in production. And we want to make sure that we have our house in order ahead of that.

Okay, appreciate that detail. Let's see, my next question is, Another one of your comments about your first contract development or development contract revenues here, and I think you mentioned you expect this to be a continuing theme going forward here. I guess I'd love to get a sense of the end market for this particular one. What's the pipeline look for this, and how much revenues would this contribute as a percentage, do you think, over the next couple of years as you ramp this up?

Hey, Carl, why don't you take that one?

Yeah. Yeah, yeah, yeah, sure. So look, this customer was in the aerospace and defense sector, and it had been a customer that we had been engaged with for quite some time. And so I think what's important is just to take a step back and really try and understand the value proposition of our product. And we talked about software definability and programmability, and a lot of customers initially take the sensor, they run it through their spaces. There's lots of things we can do on the software side to optimize it. but there are situations where customers may want to change the hardware or configure it in a certain way. A lot of times that's very easy to do just because the way the sensor is designed is very modular. You can make functional enhancements through changing out the optics lens. And so what we're going to see I think here over the next few years is we're probably going to see contract development revenues become a more meaningful number on our P&L and that's just by virtue of the fact that we have multiple different ways to generate revenues in that particular discipline. I think one is just through the software configurability piece that we talked to, but also as customers getting more involved in the product, you know, you got to imagine as well that, you know, we have customers coming to us with very sophisticated needs and different use cases and, you know, hardware changes at some point will be part of that. And, you know, That's what we're kind of flagging there. I think it's early in those stages, but we're seeing enough of the signal that indicates that this is going to become a more meaningful number over time.

Okay. I would assume with this customer and in the future as you get more of these, this would be a natural source of some volume of sensors over time here. I know that AMD customers tend to take their time going from the sort of work to volume production, but do you see these guys eventually getting to that point at some point in the next, I don't know, A few quarters, a couple of years or something like just to kind of get a sense of what you're expecting here.

I do, because, you know, the customer is putting skin in the game. And when they're paying for engineering work, they're designing you into their solution. Right. Now, that could be an unmanned ground vehicle or an unmanned aerial vehicle or something. Right. But they're putting skin in the game. They're making a bet on our product. They're investing in us. And so, you know, I think obviously it's going to be a combination of, you development revenue, but also that comes along with product sales as well. So, yeah, I think you're thinking about that the right way.

Okay, perfect. My last question, I'll jump out of line here, is just kind of looking at your business by end markets here, and you pointed out multiple times today about your success in the defense space, which is great to see here. I would assume this is your biggest market, but I'd actually love to talk and ask a longer form question to Matt here in the auto space here about the kind of the RFI, RFQ process with ones you've already been in and any future ones you're seeing here coming forward, particularly as I think you called out level three, level four programs. Thank you.

Yeah, great. I think it's a great call out, Richard. Defense is certainly making up, I can say, a majority of our pipeline and certainly most all the growth that's happened over the last quarter. But Automotive and trucking remains incredibly active. And in fact, we had a new evaluation start on that over the quarter, even from last time around. And we've had an RFQ come in. So I would say it's been steady. And the automotive programs take a lot longer time to process. So we talked about months for some of these other markets. but the automotive guys, since they are ramping volumes that require a substantial financial commitment, their evaluation period is quite a bit longer. So everything we've started this year continues. At this point, we've seen a new one come in this quarter. So steady as she goes. I mean, I think that's the way I'd say it. And it's good to see a new name come to the picture since we had the last call. but it does take time. It's a long process and the process continues. All right.

Understood. Thanks for all the detail. I will jump the line, guys. Thank you.

Thanks, Richard.

Again, if you would like to ask a question, just press power followed by the number one on your telephone keypad. And our next question comes from the line of Beau Frapp with Alliance Global Partners. Beau, please go ahead.

Hi. Good afternoon, Matt. Good afternoon, Conor. Can you expand a little bit more on the benefits of your technology that not only attracts customers but also retains them?

Yeah, thank you. I'll take that one, and good to hear your voice again there, folks. Look, there's a very different thing we're seeing that are getting folks in the door, and that's the, we'll call it the one-kilometer sensing range headline. What customers understand is that this kilometer, it's a statement about how much budget they get to spend on performance. Conor alluded to this in the script about the market shifting towards meeting their use cases rather than trying to conform to a sensor use case. The good news, when we talk about a kilometer, it's a budget, it's a really big budget that the customer gets to spend and they get to spread it across different technical vector detection range, how far they can see, what's the frame rate, what's the resolution, so on and so forth. We give that flexibility through the software to find people. And this was really essential in our collaboration with Alive3D. They had a use case that stressed a particular portion of performance that we're working with them who would spend that budget, if you will, and really push it towards their very unique and very special use case. By the way, there was some fierce competition with other players in the space, and because we were able to reallocate that performance in a very unique way, nobody else can do that, as far as we know. And that's why we won that project with them, plain and simple. So as the world begins to understand physical AI, the flexibility, we believe, is key to helping them move forward because we don't know all the use cases. We didn't know all the use cases two years ago, but we knew that we're going to see things that we didn't think of in advance, and sports analytics is one of those topics, and the fact that where we're so flexible, we were able to make the sensor meet the customer's use case instead of the other way around where the customer has to confine their use case to what the sensor can do. And that's just one data point, and I'm confident that we'll see other cases like this, a growing number in the back half of the year where our flexibility and able to allocate that very generous performance budget is going to win us additional deals and yeah, it's very exciting.

That's great. Can you give a little more color about the interest you're seeing in the defense sector and maybe just are you also seeing a broadening abuse cases there?

Yes, absolutely. So just to kind of fly back over what we talked about in the script, so number one, I think Richard touched on this a little bit, it's our largest who are the source of income at this point, our most popular segment, and not only that, but the one that grew the biggest quarter on quarter. What we had been very deeply focused on, I would say, over the last quarter was unmanned aerial vehicle, even manned aerial vehicles. We talked about collision avoidance cases like power line detection. Very important topic is unmanned vehicles, and even manned vehicles like helicopters continue to expand sticky needs and autonomy needs. And then similarly, unmanned ground vehicles. Think about it as a subset of what's happening in the robo-taxi space. Our lighter has some very unique characteristics there, for example. doing a great job at just one example, doing a great job at seeing chain link fence. Chain link fence is a very difficult problem for many sensors. And you can imagine, you know, military vehicles navigate, you know, in very rugged territory. We do a great job there. And that's just one example of where a sensor's capability has attracted, you know, part of the UGV unmanned ground vehicle sector defense. However, Q2 has been a very large influx of counter UAS, which in plain English, it's counter drone detection. And certainly, and unfortunately, there's a lot of conflict in the world, but drones are becoming a huge factor in that conflict. And LiDAR has some very unique capabilities to bring to the table in that area. And it's become very busy in the counter UAS space. over the last quarter. And by the way, defense demands performance. It's a fabulous fit for the capabilities of that.

Larry, thanks for taking my question. Sorry. Yeah, I was just going to say one thing just to add to what Matt said. You know, we may initially engage with a customer about one particular use case, and a lot of times that may lead to opportunities elsewhere. So initially we may engage based on unmanned ground vehicles. And then as we start to move forward and they realize the capabilities of the sensor itself leads to some other opportunities. And we've seen other opportunities expand more broadly into UAV counter threat detection, as Matt mentioned. So I think that's good that initially we start very narrow and we can go broad. And it's a sign that the customer is engaged and excited about the technology.

There is no further question at this time. I will now turn the call back over to Matt Fisch for closing remarks. Matt?

Hey, great. Thank you. Thanks to everybody for joining the call today. We appreciate the great questions and the dialogue that we've had here. And really excited about what's around the corner in the back half of the year. We look forward to coming back. towards the end of Q3, early Q4, and updating you on our progress. Thank you, and have a great day.

This concludes today's call. You may now disconnect.