Everyone, and welcome to Day Two of Rosenblatt's Fifth Annual Technology Summit with a focus on the age of AI. My name is Kevin Garrigan. I'm a Semiconductor Analyst at Rosenblatt, and it's my pleasure to introduce Ouster's CFO, Ken Gianella, and SVP of Strategic Finance and Treasurer, Chen Geng, for our discussion today. Ken joined Ouster about 100 days ago and previously served as CFO and COO at Quantum Corporation, as well as held multiple leadership and finance roles at energy, water, and smart city technology companies. Chen has served as SVP of Strategic Finance and Treasury at Ouster since July of 2021 and was previously Director of Corporate Development at Ouster. Ken and Chen, thanks for joining us today. Thanks for having us, Kevin. Today we're going to get a broad overview of Ouster's business and dive deeper into topics of hardware and markets and software. To begin, Ken, you have a presentation on Ouster and the world of LiDAR, so I think we should begin there. Yeah, I want to give everyone an overview who might not have met us before, just going through the forward-looking statements. You've heard a lot lately about what physical AI and what is physical AI. The way we look at it, it's really three key building blocks: perception and sensing, thinking and actuation, integration and analytics. Now, what does that mean? We look at ourselves as the leading global provider of digital LiDAR and the sensors and the software solutions that make up the physical AI world across multiple verticals. When you think about perception and sensing, that's where our LiDAR, our hardware works, operates in tandem with other sensors such as radar and cameras, both stereo and mono. You take that data and you do a sensor fusion with a software layer. We have a Perception SDK that ingests all these multiple sensors in real time, that thinks and then actuates that data and integrates it into whatever the end use case may be, whether it be a smart city use case, an industrial use case, a robotics use case, or automotive use case. That physical AI infrastructure, mixing with applications, helps drive the overall performance and adoption of our products and services with our customer base. It's not just taking a raw piece of hardware, it's really a whole system around that we provide to our customers. Our total addressable market is roughly around $70 billion. We'll go a little bit deeper into that in a second. Our financial framework is solid. We just raised over $60 million on our ATM and have over $229 million of cash and equivalents on our balance sheet as of 6/30. We have over 1,000 customers and 300 employees globally, and we've shipped over 100,000 sensors. We are the market leader in non-ADAS LiDAR, especially digital LiDAR. Globally, we beat out every other competitor within the three main sectors that are non-ADAS. What differentiates us is really our digital LiDAR. We're riding the wave of Moore’s law. If those of you are not familiar with Moore’s law, it's literally doubling the capacity every generation of both performance of the chip and lowering the overall cost. We are on our L3 chip today in our performance and range. We have an L4 coming out. When that comes out, every generation that we've gone has significantly doubled our TAM and our SAM that we go after. We're super excited about next generations. The markets that we attack today: robotics, industrial, smart infrastructure, and automotive. At the end of FY 2024, we reported our share roughly around 25% in each one. Our founders, when they had this visual strategy of going out with the digital LiDAR, a lot of folks went right to the automotive and right to the ADAS space. They saw a different greenfield in front of them. They saw the industrial, robotics, and smart infrastructure being able to spread the portfolio. It's the same base platform that we use, but we use it across all three verticals. What differentiates us within that is it's not just hardware in the SDK platform, but we also then have applications that we're building within each of these sectors. For example, in smart infrastructure, we have our Blue City platform. That's an application that helps cities and counties and states run their intersections more effectively and efficiently. We have Gemini, which works across the security and other apparatuses within here to make it more effective and efficient. This eases the adoption of our products and services, and we continue to develop other software applications for end use cases across the various marketplaces. When you look at our long-term financial framework, we are on a path to profitability. Our goal is to achieve 30%- 50% annual revenue growth, and we're feeling very confident and reiterated that on our last earnings call. Maintaining gross margins at 35%- 40%. Even though we have headwinds of tariff coming in on us right now, we still are holding to that 35%- 40% gap gross margins. The last piece is maintaining our operating expenses. Getting leverage out of our existing operations is key for profitable growth, and it's something that the company and the management team strongly believes in and gives us the conviction that we have a path forward for many years to come. Finally, when you look at the investment highlights, when you're looking at Ouster, it's really an AI solution software play. Yes, we have the LiDAR, and our digital LiDAR is the instrument and sensor and the perception box that differentiates us from everyone else. It's also adding on the AI software solutions that really differentiate Ouster from its competitors. Precision detection, classification, tracking, driving these software-based margins makes it sticky for the product for years to come. Our digital LiDAR technology on the CMOS architecture and following that Moore’s law, that’s another differentiator versus some of the analog plays that are out there. It gives us the ability to really leapfrog our generational technology from each generation that we go out. Lastly, while others are trying to figure out where their individual market spaces are, we didn’t place all of our chips on one block. We have a diversified and proven business that is winning in both robotics, industrials, and smart cities. While we do have an automotive business, having that diversification, the amount of use cases underneath those really diversifies our overall go-to-market strategy and gives us comfort. You can move and jive with each different product line or each different sector that starts to take off. That’s it for the overview, and Kevin looking to go deeper for you. Okay, thanks, Ken. I appreciate the overview. I guess just to start out, you know you're approaching your four-month mark at Ouster. What originally kind of drew you to the company, and how has the last four months kind of gone? Great question. First off, it has to be the management team. Life is too short to work with people you're not having fun with. These guys are super smart, super laser-focused on execution. Just having a really great management team around you makes the job so much easier. That's borne out to be true here for my first 100 days. Next thing I look for is the technology, either a market leader or fast follower. I can tell you right now, the tech that we have and the tech that's coming down the roadmap is super impressive. Just the vision that they had of the digital LiDAR and attacking the marketplace with this technology and being the market leader in non-ADAS is just super exciting. That's something that I found interesting. Last but not least is really the sector of physical AI and the growth associated with it. The underlying premise of making the world a better place through automation, it's something, as I grew up in industries, we called it the Internet of Things. How can you take dumb things and make them smarter? How can you take things that don't connect to the world and connect them to the world? Physical AI is taking that to the next level. It's how do you not only connect these things to the world, but how do you automate them? How do you train them to be autonomous and work for the betterment of mankind? Every sector that we operate in, it's doing something to really help and achieve that goal long term. Yeah, that's fantastic. The first topic, hardware, I think kind of on that front, Ouster has always emphasized the digital LiDAR architecture. What are really the key advantages in terms of scalability and cost versus some of the analog competitors? I think first and foremost, it's being really focused on the market validation, right? You have to understand where your sweet spot is and what customers are going to want. You have to meet the target specs for range, reliability, and safety. When you do that and you design your chip and you design your product, you're not waiting for a market to come to you. You're not waiting for things to come to you. Being able to get that volume and that scale really comes from understanding the market and working with them to validate where we're at with our customers. Once you take that design and you can scale it, that's really through gaining efficiencies with our overall capital of how we operate. We have a really good operations team. We've consolidated our manufacturing lines to, we're very open about Thailand, working both Benchmark and Fabernet. Having those costs and being able to leverage our overhead costs through growth and getting the leverage out of that is really what helps us. Plus, the natural Moore's law of being able to lower your overall cost for every new generational chip. Those things combined are really what's driving the overall profitability. Yeah, and you know, kind of going on that, on the Moore’s law portion, it has worked out very well for Ouster so far. On the L4 chip, can you just kind of remind us of the timeline on that? If you guys have set at all on what that chip will really bring to the hardware and performance? First of all, the company has learned that over the years, we don't announce products until they're ready and available for market. You're not going to catch me in that trap today, Kevin. You know, we are super excited about the progress to bring the next generation of custom chip out. As I said in my opening, each chip literally doubles our TAM and our SAM and the use cases that we can do with that, right? It's not only just leaving the use cases behind, it elevates them with more feature sets with our existing use cases, but each new generation will open up for new use cases, predominantly moving deeper into the automotive and ADAS market spaces, as well as moving deeper into the functional safety spaces that are in the industrial track. We are super excited. We believe these innovations are really going to solidify that growth trajectory that we put out there of that 30%- 50%. Yeah, Kevin, that makes sense. Looking at your hardware roadmap, just in the context more of on the AI front, are you guys seeing sensors just being designed more to offload a lot of the processing at the edge? Oh, absolutely. Adding edge capabilities and processing is almost paramount. There's so much data that comes off of these assets that you have to be able to process, think, and act all at the edge. Being able to install that, whether it be an intersection or robots working within a logistics center, these things and having that firmware enabled to work with the LiDAR sensors, that Perception SDK that I talked about earlier, to think and actuate and then be able to integrate that into the larger physical AI platform that our partners are using is key. When you think about the new features that we're rolling out in both our Gemini and our Blue City product, this provides customers more way to act on the data and really get value out of the sensor. It's not just a box sitting there generating data. We're helping them look, take action, do the analytics, rinse and repeat. Got it, got it. Yeah, we're going to talk about software in a little bit, but I guess just to kind of delve into the four main verticals that you guys are in, smart infrastructure, robotics, industrial, and automotive, which ones are you most excited about over the next 6- 12 months? Is there a vertical that you see inflecting first in terms of AI adoption? What vertical do you see kind of driving higher adoption of LiDAR sensors? You're not going to like my answer. The answer is all of them. You're not going to make me do a Sophie's Choice here the first 100 days on the job. What really attracted me as I talked at the beginning was we did not just pick one vertical or one use case to bet the company on. We're spread across four unique verticals that all have unique video, our unique use cases to them. I'll start with the one that's really impressing me just because I come from a smart city background, which is the smart infrastructure. If you look at the performance benefits that folks can get out of that, we have the end-to-end platform and solution all the way down to the analytics that can adopt and really work for the cities and these intersections and get the accuracy and the multi-object identification, larger coverage areas than the incumbents are using to run these intersections and give them data to make intersections safer and more reliable for traffic management, pedestrian management, etc. It's super exciting and really smart infrastructure is growing faster than any one of the segments, predominantly driven from the applications that we have there. Industrial robotics is really a cool area. If you think about the robotics side, we just made the announcement about Blue UAS coming out. I think we'll talk a little bit more about that. If you think about just getting into drones and being the only Western digital LiDAR company, let me take that back. We're the only LiDAR company that has this certification out there today. Being able to work with the Department of Defense and having that certification greenlight our supply chain to be friendly and safe for manufacturers to use, that was really a great check the box for us. I think that's going to open up new worlds for us. Industrial is also an area that I'm super excited about because the industrial side, think of like the logistics and warehousing and some of the other elements there. There's huge brownfields that we can go capture that some 2D LiDAR companies are operating in today. People already understand the benefits and efficiency that LiDAR can bring to it. Now add on our digital 3D LiDAR and our backwards compatible software that can work with multiple generations. That just opens up a whole new field, brownfield of opportunity for us as the new generations come out. Last but not least is automotive, the new generation coming out. That's just going to be super exciting. It's going to be really expanding us deeper into that automotive marketplace with ADAS and being able to go head-to-head with some of the folks that started in that area. We're going to grow into that. That makes sense. Kind of delving into each of the four verticals, I mean, with AI really enhancing perception for mapping and infrastructure, how do you guys differentiate yourself against the competition, which I believe is really just kind of legacy camera players? Yeah, you know, I think the biggest thing is the software and the technology, right? If you think about the perception of what our, when you look at the perception and sensing realm where you operate in, we're at the higher end than cameras and radar. The ability to see when the sun is shining, when it's foggy out, the ability to work at night through all different environments, our equipment is rated for both super hot and super cold environments. That ubiquitousness of the technology is a key differentiator, let alone the granularity. I mean, you can see down to a centimeter with the LiDAR that we put out there. You combine that with the software elements of it, and it really changes the dynamic of what we can bring to the table. We work in tandem, right? I don't want to say anything negative about any of the other parts of perception because I think, you know, if you look at the stereo and the mono base cameras that are out there and our LiDAR, we see them hand in hand. The reason why I say that's important is because there is a spectrum of what customers need in ordering this perception and sensing. Where we want to really differentiate ourselves isn't just on the hardware piece of it, but it's having that Perception SDK and doing that sensor fusion. Being able to make the LiDAR and the camera work together in tandem and process that in real time so that you can think and then, you know, take action. That's really where the long-term vision is because as you get past the first and second movers within physical AI space, you have a long tail of fast followers and laggards out there that are going to be looking and needing help. They're not going to have the expertise or the software design to be able to get their programs up and running quickly. By us having these SDKs and the hardware all bundled together for a perception system for them, we'll really differentiate us in the long term. That makes sense. I think it helps investors kind of frame how big of an opportunity a lot of these markets are for you guys. Can you just kind of, I know you said it's a $70 billion TAM, but just an example of how many intersections you guys potentially could be in? I know you've said it in the past, so I think it's about $300,000. Yeah, over $330,000. Yeah, and if you take those intersections and we're in the hundreds right now of that opportunity, and our product is price point well against the competitors, but the feature functionality far exceeds anything that our competitors can do. Taking not only just that intersection as an example of the early days that we have there, the announcement we just made with Utah is a great example of that. When people get our technology and they can see what it can bring to the actuation and the data and analytics that come off it to provide deeper safety to their city, you're looking at that at the exact same price that you're supplementing, not replacing humans, but you're using the physical AI part of this to learn about what happens on traffic patterns, time of day. You're learning and using the Blue City software in order to control and be more effective and efficient with managing intersections. All these things are just force multipliers for our product versus some of the existing competitors out there. Moving on to the industrial and robotics end market. It seems like those are really among one of the fastest areas of adopters of AI. I mean, how are you looking at those two markets? How are you capturing the demand? You mentioned before about certifications. How do the certifications help you guys in those end markets? Let me first talk about what's really cool about it. Let's start with industrial. You know, if anyone's seen the movie Interstellar, there's a scene with Matthew McConaughey where all the, you know, Caterpillar tractors, all the tractors come to his house and work autonomously. That's what we're building today. We're working with partners in ways that you can automate tractors, you can automate mining equipment, things that require skilled labor to go and run. Think mining, you know, in the outbacks of Australia where you literally have to create towns for people to come and run these vehicles, or middle America where you have vast fields and you need qualified people to work and run to harvest the fields. Having these systems be autonomous and run and do the basic chores or have one person run four or five giant rigs at one time and let the AI do it for you. These are use cases that we are in right now today, and our partners are prototyping. You take that and you say, okay, great, now that I have that, let me process it through and what does it look like in the warehouse? There are several big companies out there running. I always say the world comes to an end when I can't get my packages within two business days. That's just the norm for us now. That takes people to make it happen. With labor shortages, it's really important to have AI and the autonomy to operate within the warehouses, to do the picking, to do the moving, and get it to where the products need to be. Our technology is part of a system for perception and sensing that helps these people work more efficiently and effectively. If you just use those two use cases that I just described of rigid, large outdoor rigs and then in-house warehouse logistics, those marketplaces are huge globally as the automation takes over to displace or supplement, I should say, not just displace, but supplement the manpower that's needed to operate in today's world. Yeah, there's a lot of runway in those two years as well. It's unbelievably early innings, Kevin. If you think about just where we're at in that prototyping that we're doing, for example, there's over 2 million tractors produced a year. 2 million, right? You just think about that's not just replacing the old ones, it's these new ones being sold in. As you go through this life cycle, that's a really long tail of once you start doing this automation, even if it's not self-driving, just for the safety of it, right? For people operating, you could have different levels of automation to operate these. It's a super exciting landscape. I can imagine. Last but not least, the lovely automotive market. How do you guys see ADAS and autonomy adoption timelines kind of evolving? You guys do have about, I think it's about 25% of your revenue comes from automotive. What parts of the automotive market are that, and where do you guys have this kind of the strongest positioning? Yeah, today we're focusing on Robotaxi and through the shuttle side of it. We're working up to the commercial ADAS with our next generation. You know, we're super excited about opening that market up. I think a lot of our peers who came out of the gate several years ago and just focused purely on that. I think the thing that I want to use here real quick, Kevin, is, you know, when I came into this, I heard a lot of investors talking about, oh, LiDAR, it's all the same. Someone who uses this can be the same here. It's not. It's a category, it's a technology, but the reality is every LiDAR is different. Whether it be the beam wavelength, whether it be the killer of the beam, whether it be the receptacle, we talk about our CMOS-based digital LiDAR architecture strategy versus an analog strategy. What you have a lot of these folks within the ADAS, they have a product that shoots really far ahead, a single beam to look at things 300 m, 400 me down the stretch. We took the approach of we started up close. We can see zero out, and then we've been building, and we have technology that goes up to over 200 m with our short and medium range. We also are the only ones that have a 360-degree view. If you think about how we operate within the robotics, the industrial, the smart infrastructure, our technology is bespoke to this genre of being able to hit the technology for what they need, where if you're trying to jump from just a single beam that shoots really far straight ahead but can't see up close, it's really hard to enter these other marketplaces. As our technology is going, we already have the 360, we already have the up close, it's now focusing and narrowing that beam to look farther. That was a strategic decision that our founders made over 10 years ago to say we'll slowly move into that marketplace. I think that when you think about the automotive market, it's still ripe and new. It's not moving as fast as people anticipated. You're not seeing the adoption on the commercial side as quickly as people want. I think this gives us a great opportunity as our next generation comes out, to capture either share that's already been awarded to other people or come in and grab greenfield or that second followers and the fast followers that will follow on from the first movers that have already adopted it. I still think it's a real fertile space for us. I don't think we've missed the boat at all with it. I just think we've methodically moved our way there. Yeah, the market's keeping growing. I just said everyone always thinks about LiDAR in the automotive market just about long distance. You know, how far can I see? It's a lot different, especially when you see some of these Waymo taxis that have several LiDARs because you need to see up close. It's a lot different when you don't have a driver in the car. Yeah, I think it's a great proof point for the sector and industry and autonomy overall, right? You got to get it right. You don't want to hurt somebody in any way. I think that's one of the reasons why, back to your product question, we're so thoughtful about our releases, right? You have to be positive that you will be able to actuate a braking motion for a vehicle if something comes in. You got to make sure that you're not going to hurt somebody inadvertently because of your product in a warehouse or on a mining site. You have to think holistically. This isn't about just throwing tech out there. What's different about physical AI is these are physical things operating in our world that you have to be thoughtful about the technology and the software that you're putting out there. That creates almost a natural moat for what we do in this sector, in this industry, because it's really hard. For companies like Ouster who have really perfected this and have honed in on this and focusing, like I said, on these other areas and moving into automotive, there's a lot of learnings here that we've taken. With the combination of Velodyne, our birthright is LiDAR, right? They hold the foundational patents to LiDAR. If you look at the company as a whole combined, we have a really strong patent portfolio combined with really cool tech, combined with cutting-edge AI software. Put that all together, you have a winning perception and sensing physical AI system. Yeah, again, I think to your point, a lot of people miss the safety aspect of what LiDAR sensors can bring. Everyone thinks that the tech is awesome, but it's a safety mechanism, which is huge. Okay, great. Moving on to kind of the software, you know, where do you see monetization opportunities? You talked about AI-based perception. How material do you think software could become relative to kind of hardware sales? In the long term, I think it will be material. In the near term, probably not so much. I'll tell you why. The strategy within this sector right now and more physical AI, it's new within the genre of what we're operating in. If you think of the larger industrials, robotics, et cetera, the vision here for the company is land, win the node, win the tech with whatever we're going after in each of our sectors, expand that into production with the companies we're working with after you win the node. The underlying software layer that's supporting them, you can upsell that with applications, feature functionality, and customized support. Ultimately, in the long run, as I said, you have a lot of first movers who want to design it all themselves or they come with a vision and they want just your SDK and your hardware to plug into their design. That's great. We're here to support them. Early first movers, not so much opportunity there. As you start getting into production with them, they're like, oh, I really want that feature functionality. Oh, I need this support for it. You start seeing opportunities to monetize the software as you expand out, and then you can uplift up the software sale. Naturally, the people who are second and follow-on movers, they won't always have that expertise or they're looking to catch up. They're looking to monetize and get to market quicker. We fully expect them to buy the whole turnkey package from us for that whole physical AI for perception and sensing all the way through thinking, acquisition, and application that we can bring that to market for them, either directly or through partners of ours that plug into our APIs. I'll be excited when you guys break out the revenue line item for software. I'm going to ask you guys every quarter when that's coming. Our goal right now is win the node. Right? It's very important in these early days that you're winning the footprint, and then once you win the footprint, you can expand and upsell from there. If you're not winning the footprint, that's the reason why we focus on the verticals we do, because you never know which one is going to break out quicker or faster. This gives us the opportunity to look across the portfolio and have opportunities to expand quicker because as people go into production. Yeah. It almost sounds like you guys are creating a library similar to how NVIDIA has their CUDA libraries. Can you just kind of talk about the different software strategies by end market? Is it difficult to apply, you know, one software application to another end market? I don't know if that question kind of makes sense. No, it does. What's the portability of it is the word I'll use. I think Gemini is probably the best version of portability when you think about how it can operate across multiple sectors. We're really proud of what the Gemini software product can do. I actually work here in the San Francisco office and I've embedded myself around the Gemini part of the Gemini team and the hardware and software teams here. Super, super smart people working on critical problems and critical end use cases. I think Gemini is a great example of how we're looking at that to expand across multiple verticals with Gemini+. It's like plus logistics, plus this, plus that. I think a great piece of that is one of the acquisitions that we made early on and came in with the Velodyne purchase was Blue City. If you think about what Blue City can just really is turnkey for folks to plug and play and get their intersections up and running, that's a great example of where, okay, that's more feature purpose or application end case purpose. Being able to, it still builds off the same Perception SDK that we have up top. I think that's the takeaway I want people to get from this is that once you have a really good sensor fusion platform that can pull all the data in and manipulate it real time, how you feed that into the applications and the end use cases, that becomes the next step within the process. We got Chen, myself, and the team are always looking for good opportunities out there, but also our partner network. We have a lot of folks that come and use our technology and then are building cool end cases. I think having that ecosystem, Kevin, is what really differentiates us. We're not afraid of an ecosystem. We want to encourage people to build across upon our platform to come up with new use cases because it just makes our tech better and more in need the more use cases you can add to it. Delving a little deeper on that and, you know, customer adoption, can you just give a couple of examples of where, you know, the software has materially increased, you know, the value to customers in the sense that, you know, you talked about Blue City? If you just buy the hardware, you just get the point cloud, but if you buy everything, then you get, you know, the data analytics part. Where can you just talk more about kind of what customers have gotten in terms of value? Oh, I mean, it's the time to value, it's speed to market, it's the realization of, you know, the service that you're buying, right? Being able to set up a Blue City or using Gemini for crowd analytics, you know, as we're doing across multiple sites, for stadiums and crowd control. These are examples of where on their own, these cities probably wouldn't be able to realize the value quickly. Cities operate the same way as business. Show me what I'm getting for my money. Being able to plug and play and have the analytics spitting out and seeing real time what's going on in intersections and helping them plan and use the analytics to manage their city day one, that's super important. That time to value is really what we bring to our customer base with an example like Blue City. Yeah, that makes sense. Okay, perfect. I guess this is the last topic that I wanted to hit, and we got about 10 minutes left, just on the financial front. You talked earlier about the path to profitability targets. Can you just go over those again and what kind of gives you guys the confidence that you can hit those targets over the next few years? Yeah, I'll give the three real quick, just to remind everyone. You know, from a growth perspective, we're saying 30%- 50% growth. I'll unpack that in a second. From a gross margin perspective, that's 35%- 40% in gross margins. This quarter, we ended at 45% gross margins. We had a little bit of a tailwind from some tax refunds that we got in. Then, you know, the third one is just controlled and disciplined OpEx spending, maintaining our spend levels, you know, with inflation built into it, you know, similar to levels from Q3 of 2023. First, to unpack why we believe in the growth, goes back to why I'm here. Think physical AI and the end market associated with it. Each one of these three, forget automotive for a second, each one of the three, robotics, industrials, smart infrastructure, all three of them are growing at really, really fast clips. With our new technology coming in, we can do more and more from that total addressable market. It becomes a serviceable addressable market really quickly for us over the next three years. It's expanding our availability and expanding our marketplace in an expanding marketplace. I'm not even talking about the brownfields that we can go over and capture from people using this already in the industrial space. This is just pure new market opportunities. I think the second one is our technology, that technology roadmap that we're lifting out. You know, we've shown that the ability to look at ASPs, maintain and in some cases grow our ASPs over each generation, but at the same time use software and services to bolster up those ASPs as we go forward. You know, that's going to help that growth engine. When you look at the gross margin side of things, you know, just a few years ago, we were negative margins. The operational efficiency and leverage that we're getting, we shipped the most sensors ever in the company history, 5,500 this past quarter. Our ability to operationalize and scale and get leverage from the base platform that we have, that will only improve as we ship, you know, from 5,000 to 10,000 to 20,000 sensors a quarter. You're just going to get the exponential leverage out of that. Now, we do have some headwinds with tariffs that, you know, the reason why we're reiterating a 35%- 40% is, you know, we have to absorb some of the margin impact. While we have great partners and customers and we can pass some of that on to them, there's still going to be some absorption we're going to have to deal with in the near term as long as those exist. Finally, on the OpEx, it really comes to efficiencies, right? I think about any maturity life cycle of a company, there's a lot of money you're spending in areas for growth, a lot of money you're spending on consultants and inefficiencies of getting things done. The company has really done an awesome job of putting in systems, tools, processes, and people to make ourselves more effective and efficient. I think we're going to continue to see efficiency gains out of the existing OpEx space that we have and have enough go-to-market dry powder to really be strategic and enter these marketplaces and win these marketplaces. I can tell you while we're being disciplined with the spend, we are not sacrificing our go-to-market strategy. We feel we got the right team, the right set of partners, and making sure that we're going and capturing and winning. As I said earlier, you got to win these nodes, win the footprint. We're spending the dollars to go do that. At the same time, we're growing profitably. If you look at the EBITDA, our EBITDA is continuing to improve quarter- over- quarter, year-o ver- year. We think that's a trend that's going to continue. I think you've modeled it somewhere in 2027 that we end up on a positive operating free cash flow and a positive EBITDA. I think, you know, when you do the numbers, that's kind of where you end up. Yeah, it's great to see you guys. I'll switch to the gross margin front because, again, as you said, a lot of the companies who went the automotive route are either just coming out of being gross margin positive or still in the negatives. You know. I want to highlight one thing on that because it's really important back to the technology thing of what I said. The technology and the tools that we use within robotics and industrial and the use cases and the demands that these companies are looking for. If you remember when we talked earlier about making sure that there's market validation, the ASPs and the software and the services for these markets, these use cases can hold up a lot better than a single point product that's pointing 400 m ahead, where, you know, I think everyone knows the ADAS market, the price points and to get to that commercialization is precipitously low, right? You can't compare what we do in our other sectors because it's a different type of technology when you get there. When we get there with our Kronos Technology, that's purpose-built in mind and understanding there would be a different set of ASPs and a different market to capture versus what we're doing with our product sets and technology today. For those of you out there trying to mix the two together, Ouster is in a different class and category than just what you may be seeing from price points and ASPs today in the ADAS sector. Yeah, you are seeing a lot of the companies trying to shift more into the non-auto market where you guys have a strong foothold. It was a good move by Angus and Mark back when they first started the company. We only have a couple of minutes left, so I'll just leave it open to see if there's any last minute topics that you wanted to address or any last words you want to kind of leave with investors. No, we totally appreciate it. This is a leader of physical AI and the only Western LiDAR company that's in an unbelievably strong financial and market position. We're really excited about the technology and the long-term profitability of the company. Please reach out to Kevin, reach out to Jim, Chen, or myself if any of you guys have any follow-up questions or want to learn more about Ouster. Perfect. Ken, Chen, Jim, we appreciate you guys taking the time to speak with us today. Thanks.