Thank you for joining us for this session of the WTR Insights Conference featuring Circus SE. I'm James Kisner, Managing Director of Technology here at Water Tower Research, and I'll be your host today. We're pleased to be joined by Nikolas Bullwinkel, Founder and CEO of Circus SE. Welcome, and thank you for being here, Nikolas.
Thanks.
All right. Before we begin, please note that Circus's financial disclaimers can be found in the financial reports in the Investor Relations tab on their website. Also, this fireside chat may not be reproduced, nor may a written transcript be distributed without the express written consent of Water Tower Research. Investors interested in scheduling a meeting with Circus SE can indicate that interest via the conference portal. Now with those housekeeping items covered, let's jump in. All right. For those of you who caught our recent WTR Small-Cap Spotlight podcast, we covered the broad overview of Circus. Today we aim to go a little bit deeper. Circus SE is a Munich-based, Xetra-listed AI robotics company developing autonomous meal production systems for commercial and institutional food service settings. Its flagship, CA-1, combines robotics with Circus OS to automate much of the meal preparation workflow.
The company is extending that platform into defense applications with their CA-M. Nikolas, let's start with the big picture for new listeners. Circus positions CA-1 as an autonomous meal production system for commercial and institutional settings. For listeners that are new to the story, what does the system actually do today, and where does it best fit in a customer's food service workflow? More importantly, what problem are you fundamentally solving that couldn't be solved before?
Yeah. Maybe starting a bit with a bigger picture on the food service industry, right? When you look at food service globally, it's one of the biggest industries in the world, but at the same time, it's one of the least automated industries, right? There's great graphs that show how much automation, for example, pharma or automotive have already experienced over the past decades. When you look at food service, it has a similar size, a similar level of workers, but there's a very low single-digit percentage of automization that has happened over the last decades. We aim to address exactly that, specifically because right now is the perfect time to do that. On one side, robotics and AI are at a point where it's ready to scale to more industrial use cases.
At the same time, the whole industry is facing really inevitable changes, right? You have a constant rise in labor cost, labor availability, and also the level of training is constantly shrinking. It's a perfect time for us to jump into that industry, with an entirely new generation of autonomy, basically. Maybe quickly, what we're not doing is, we're not targeting the small restaurant owner here, right? It's not about bringing a robot to your favorite Italian restaurant around the corner. It's all about addressing really infrastructure food providers.
What we're targeting is everything from defense application, gas stations, a lot of cases in care, when you talk about, for example, elderly care homes, hospitals, up to gas stations, airports, retail, employee catering for factories, where it's also about the night shifts, the weekend shifts, everything where you don't really have a proper setup that can run 24/7 today. That's where we bring fully autonomous systems to the market and are solving all the mentioned issues with one single system, so food supply becomes really easy and, frankly speaking, idiot-proof to manage, without any need for really skilled labor on ground 24/7.
Okay. That's really a great overview. Thank you. You've got a couple of live deployments in Germany, including Meta in Munich and REWE's Fresh & Smart concept in Düsseldorf. What have you learned from those first live sites?
Yeah. It's kind of a fraction of the things we do, right? We deployed the very first robots back in 2021, which was a very early phase and kind of the beta madness of building entire new technology, and since then really evolved the technology right now towards fourth generation, the first generation that we also right now took towards high-volume production. That's really where we learned a lot along the way through being really the first company globally to industrialize that kind of fully autonomous systems, which also gives us a very strong USP right now to roll out to, for example, the customers you mentioned, which are all customers that tend to not work with companies and new technologies in early stage. They want to have a certain level of proof of scalability of that technology. That is something that we can provide today.
Over the past month, honestly, what we learned at the very core is that the consumer is ready for it. The biggest question mark over the last years was always how far are we right now in adoption from the end consumer that still needs to press a button to get his or hers fresh meal out. We really learned that customers honestly have been just waiting for that. You get a better meal at a better price, at a higher level of availability. Today, you get a high availability and consistency at a good price at McDonald's maybe, but not really for a real balanced and fresh meal. That is something that is really outstanding for across all use cases.
Obviously we improved on a ton of small things, not only on the hardware side right now over the past quarters, but also entirely on the ecosystem that we're creating, where the core moat is not just the hardware itself, but it's really how good are you in the overall integration with the customers on ingredient supply, service maintenance, the software layer as well. When you think about, I don't know, a Tesla example, Tesla didn't necessarily win in the early stages just because they had a great car, but because they understood that charging is the main breaking point in the adoption of the consumers.
It's a bit similar here where for us it's also about the entire supply chain integration and optimization, so that all you need to do is basically plug in the side with ingredients, close the doors, and the rest is our magic end to end.
Very helpful. I was hoping to maybe dive into one of those customers a little bit. REWE is one of several kind of lighthouse customers that gets investors' attention. If perhaps you could talk a little bit about that customer for those, particularly in the US, who just may not be familiar with the retailer, and then kind of walk us through the status of that relationship specifically and any update on how that particular deployment is going, what the expansion roadmap might look like, and what may need to happen before you can broaden that rollout?
Yeah. REWE, to give a bit of an understanding, is kind of the European REWE in a way. Thousands of locations across multiple European countries. Largest retailer in Germany as well, with a couple of 1,000 of supermarkets where we right now, across multiple cities, have integrated our systems. You really have, for the first time ever, a fully autonomous touch point, fully integrated into retail. You walk into the supermarket, you can select your meal, and you get it freshly cooked by robot and can take it out within a good three-four minutes. That is not necessarily, from my end, the most exciting case right now, right?
There's a ton of things that are happening, a ton of cases, but it's definitely the case that brings the biggest buzz and the highest level of visibility, where the core goal in the first phase was to have a case and a customer which is accessible really to everyone, and where we can have a touch point with nearly everyone across the whole society, different target groups, different age groups, different levels of income, right? Everybody has to go to a supermarket. That really gave us the core proof of the adoption. Right now it's going exceptionally great. We just extended the menu, extending and growing on a weekly basis when it comes to stability, integration, as well as the overall sales volume, and are right now advancing also across multiple retail cases.
That's helpful. It's probably one of the places where you can most easily kind of go try the product or see it in action too, if you're not working at Facebook.
Always.
One thing we talked about in our last discussion, you mentioned that there's some scaling required to build a reliable ingredient and supply infrastructure alongside the kind of core robotics technology. Just where are you sitting on that front today? How scalable is the system now across multiple sites?
Yeah. As mentioned a bit, right, the core here is building an entire ecosystem. We are kind of flipping the entire structure of an entire industry, right? What you have today is you have central production kitchens in the middle of nowhere, which pre-produce meals, and then send them kind of pre-packed out to, for example, a supermarket or to a different canteen setup, which always leads to a high level of waste and a very low level of quality. We kind of flip that, and we're bringing the factory directly to the end consumer. It's all about the ecosystem. What we did there is, we also built the entire stack of AI fine-tuning models in the background, plus an entire software layer based on the Circus OS, so that we can control the end-to-end value chain.
That's also something that really makes us stand out when it comes to really winning the big customers, that we can, for example, right now control every element of, also from a regulatory point, required hygiene protocol, temperature protocols, the entire structure of buying, selling the meals, prediction elements of food waste, for example, to optimize on your margin profiles when it comes to the procurement process, as well as the offering for the end consumer on the menu side. That's really right now one of the biggest factors also that we're investing in for future growth along the entire supply chain of food procurement.
Great. Thanks for that. Let's turn to technology for a bit. You described the proprietary Circus OS as the backbone of the entire ecosystem, powering recipes and procurement, hygiene compliance, and more. Beyond just controlling the robot, how does the OS create a competitive moat? As you add more units to the field, maybe talk about the data flywheel. What are you learning from those 2,000 or 3,000 data points per meal, and how does that sort of translate into a compounding advantage over time?
Yeah. When you look into food service canteen operations today, it's a pen and paper business, right? Every chef knows a little bit, and then there's one person who is doing the procurement, and everything is a bit based on a rule of thumb in a lot of cases. That goes for big quick service restaurant chains as well as canteen operations. The lack of data leads to a lack of margin because you lose your raw efficiency, because you really don't know what is being sold. What do I still have in stock? Based on the ingredients that I still have, what kind of recipe can I offer, right? There's a lot of things that come with it. What we are doing is basically taking all these data points as really the first company ever globally who is even able to do that.
McDonald's knows a lot of things about burgers and fries, but really when it comes to the huge variation of meals and different culinary target audiences, then it becomes very tricky. That's something that we've built over the last years. Beyond the software and the hardware layer, there's always an entire intelligent AI layer where we bring all the data together and get very smart or drive very smart decision and automated decisions when it comes to the supply chain process. That's one. The second element is also the AI-based control system of the hardware itself. It's not a humanoid robot. We don't have the legs, and we don't need the legs for cooking. When it comes to the arms and the overall handling of the machine, that's where we brought in a lot of intelligence.
I give you one very plain example, and every human knows that you don't move a soup in a pot as fast as you would move a salad, right? You would spill the soup across the whole room. A robot arm by default doesn't know that. We brought a lot of intelligence in right now with multiple cameras and our own vision intelligence layer to, for example, control these things, right? Liquids, movement patterns, as well as predictive maintenance elements so that we can understand if, for example, a robot arm is about to break in six-month time, we can already realize that in early-stage patterns of changes and movements when it comes to milliseconds. We use really our AI layer across the whole ecosystem for a high level of stability and a better level of forecasting.
Helpful. This is kind of related point. Just looking at the kind of autonomous kitchen space today, admittedly, it's pretty early. Who do you see as the closest competitors, whether it's other robotic startups or the traditional contract caterers like Aramark or Compass Group or even larger tech companies that might enter the space? Who do you think about? Is there anybody that could offer serious competition at all? What's the competitive landscape technology-wise?
Yeah. There's a good handful of players in the market globally which develop technologies in the same direction as we do in an earlier prototyping stage, but still. We're always very happy and grateful about every new peer who's entering the market because, as you mentioned, it's quite early. We are always more in a mode of educating the customer and our B2B customers as well to showcase that this technology is scalable today, that it tastes well. That all helps. When you look into the field of where we're taking business or who are we taking business away from, it is definitely the large-scale caterers.
It's an Aramark, it's a Compass, it's the Dussmann of the world, which do kind of the same thing we do, but they do it with 20 people on hundreds of sq m, and we do it on a trailer-sized robot with zero people. That's really an entire new level, entire new generation of services where you don't have that kind of level of fixed cost and dependency on humans that you don't even always have and that are not even always available. I think there will be a massive shift in the entire industry towards autonomous solutions, and I think it's the naturally next big industry that will advance towards full autonomy after automotive and other higher or high-tech industries. There will be a big shift when it comes to caterers versus what we can do.
I want to turn to one final topic here in technology, and that's agentic AI. That's become one of the hottest themes in the market discussion of AI pretty rapidly of late. You guys moved relatively early, I think, with the Fully AI acquisition. As your product has matured, how has your view of where agentic AI adds real value inside autonomous food operations changed, perhaps over the last six, nine months?
Yeah. In software it's a radical dynamic of change right now, right? With a lot of people saying, like, software is dead, it's all vibe coded right now, and like SaaS is going through a radical change, how we see it is, and when we also speak to our clients is, it's not that they will right now build the entire complexity of management and intelligence themselves. What is changing is the interface of software. The future of software doesn't look like software is looking today anymore. You don't have a very complex interface with 1,000s of tabs and buttons, so you need to manage everything yourself. What we are building is a lot more of a proactive layer for managing your operations. Through the agentic layer that we build in, we give you a recommendation of what ingredients to buy.
We tell you when you need to interact with a robot, so you don't need to always wait for the next alert or try to figure things out yourself through the data that you're getting. We give you the recommendation and kind of the chatbot layer. You always have a very skilled chef assistant next to you, which is completely based on the data that we're getting. The whole management of the new technology becomes very natural and easy without the complex software system.
That's helpful. All right. Let's turn to defense for a minute, another big growth vector for you. You kind of made visible progress in defense, got engagements with Ukrainian partners, and then the German Armed Forces as well as NATO and US certification. How far along are you today in moving from initial deployments into kind of structured procurement decisions and more discussions, and what milestones should investors watch for next in this area?
Yeah. I think honestly, it's a growth market for us right now that surprises me every day. When we entered the market right now and also with the first contracts over like Q3, Q4 last year, the expectation was always like, okay, it takes a good one, two years to really get the product live, right? Integration with armed forces is quite complex, also complex when it comes to cybersecurity, integration to their systems. It's a different market compared to a commercial rollout. We're moving so much faster than we initially had anticipated. As we speak, we're ramping up the first units, for example, with the German Armed Forces, which I think usually with a lot of different cases took years. Right now, the speed of deployment really accelerated over the last years when it comes to new technology integration for all kinds of armed forces.
We are really benefiting from that. We're very much unique with our technology in that market. When you think about drones, there's a ton of right now players in the market where it becomes or has become quite competitive. When you look into our systems of really autonomous ground infrastructure, that's something that is right now really at the early stage of completely ramping up because all the armed forces realized one level is the weapon systems, but the other level is the ground level. Where the armed forces spend the most money today is labor, but not labor in combat. It's labor in logistics and base infrastructure tasks. That's where we are winning right now, a ton of customers, and where we're growing actually the fastest.
Looking into the growth on the defense side over the next years, I'm very confident that we will get to a point where this will become a good 50% of the revenue over the next years quite rapidly, and where we're growing also faster than expected. With having, for example, the German Armed Forces, multiple cases in Ukraine and a lot more names to be added to that list in our releases upcoming over the next few months, that's really where the growth is for us right now.
Great. The CA-M is a 20-foot container of a system with a lot higher throughput than the CA-1. I think it's 300-350 meals per hour versus 80-100, roughly. Just in general, how should investors think about the CA-1 versus CA-M opportunity in defense?
Yeah. We're having two systems today in our portfolio, with obviously more to come. One is the smaller CA-1 system, for example, integrate into retail, which is not excluded from defense. In different defense operations, what you want is something that is flexible, mobile, and also can be used in outdoor environment, which brings a different, less stable setup when it comes to the base infrastructure. We developed a system for that, which is containerized, in a ton of case will be bulletproof, right, so that you can really have a sustainable and lasting structure for meal supply for soldiers in that case. With the CA-M, we can offer exactly that without the need for having multiple people being allocated to meal supply for different troops on site. There's a ton of more use cases for that system.
It goes far beyond defense and food operations as a core.
Actually, along those lines, so I think you mentioned construction sites, disaster relief, outdoor events. I guess maybe you can give us a little color on kind of where CA-M is getting the most traction.
Yeah
Military applications.
Right now, we're completely focusing on the military application. That's keeping us busy right now for the very moment. We get a lot of requests right now from different setups. Construction sites will be something that is next. We also actually don't even have an infrastructure yet that's about to be built, right? It's a perfect integration case for us, and a very similar setup kind of to defense operations. It goes, as you mentioned, right? A couple of examples, festivals is actually something that is heavily requested. From Burning Man to Tomorrowland, there's a ton of conversations that we already had to also have a more reliable 24/7 infrastructure without relying on labor in these camp scenarios.
I've heard that the environment in Burning Man can be pretty harsh sometimes.
Yeah.
It makes a ton of sense. That'd be very interesting to see. Let's turn to financials. You already gave 2026 guidance for EUR 44 million-EUR 55 million in revenue, an EBITDA loss of EUR 6 million-EUR 8 million. It'd be a pretty big inflection point, kind of your first big commercial year. One thing that's interesting is you kind of described that as maybe conservative, just probably because you're not assuming kind of new customer wins in there. I am just kind of wondering what would happen, what would have to go right for you to land in the upper end there, and are there any significant risks, whether it's kind of manufacturing, scaling? There's always risks in every guidance we know.
Yeah
Anything you kind of call out?
Yeah. I think when it comes to the different risks that we see today in the market, production risk itself is the smallest risk, right? It's really more about having the entire infrastructure and supply chain layer, having the right conditions on the site. For example, in the Ukraine, the complex part is not really putting the hardware, it's the supply chain, right, making sure that there's a constant flow of ingredients, right, so the whole management process behind. That's where always the whole intelligence and software layer kicks in, plus the integration layer of things that we are really deeply connected with their systems and can get the data and deploy the intelligence layer. When you think and look into this year, it's not the easiest environment to scale when it comes also to different geopolitical scenarios, right?
That doesn't necessarily make things easier, but it also doesn't touch us right now radically in our deployment speed, where it's more of integration. What we'll definitely see over the course of the year is a structure where also the variation of products will be more reflected in our portfolio. We initially planned the guidance with a core focus on the CA-1. Right now, we see a lot more growth on the CA-M. We have more products coming out to the market this year. The balance of the overall revenue split between different products in the portfolio will grow. There is also obviously more growth scenarios. Right now, it's a bit limited how much I can comment on the guidance right now. Overall, we're quite comfortable. We are on track in Q1 with the deployments, and it's great what we see in terms of our progress.
Good to hear. Let's talk about your recent capital raise. I think you raised EUR 30 million in December 2025, and both you and your CFO participated, so that's always good to see. How do you expect to deploy those proceeds, but also just what do you think your cash runway looks like today? Do you have thoughts on when you might expect to reach cash flow breakeven?
Yeah. With the last funding round, it's multiple things, right? One thing is really speeding up production and also the procurement time. What we've seen in the first unit batches of the production of the CA-1 is that the majority of time was lost in the procurement process. For example, sensors right now is highly requested in the robotics field. What we're right now doing is we're leveraging a portion of the proceeds also to become a lot faster in technical procurement. With that, production becomes faster. We're not pre-financing entire hardware elements and taking our own risk on the books, but when it comes to certain components, that just accelerates production. Where we produced over the past year, a few units in six-eight weeks' time, right now, we cut that in half, right?
We are becoming a lot faster over a short amount of time, which will right now also be the visible part in accelerating the deployments over the course of the year. That's one. Second is obviously really keeping our leading and pioneering position in the global market. There is a lot more of technology that will get to the market. There's also a ton more competitors growing, right? When you look into the US market, there's Marc Lore with Wonder, who's doing amazing things. I just saw that, for example, also Travis Kalanick right now with Atoms, is also looking into that field. Autonomy in the food service space is really a hot topic and becoming a hot topic. For us, it's really about growing our IP base, growing the structure of the patent technology that we have today into more use cases.
Also, a large portion of the proceeds goes into R&D of future products and obviously rollout-related positions, headcount, and so on, that we can really grow faster. On the profitability level, right now we don't really have a ton of costs, right? We're not 1,000 people company. We're right now running with less than 100 people, and probably right now for the next month with the exact same amount of AI bots. Like the amount of push that we're right now getting through the integration of AI code and AI in overall processes is simply amazing, and the output for a small company is really growing exponentially, through just integrating that into every step of the company. We don't really expect a large amount of pure human headcount growth.
Really growing also through a high level of AI output, and driven by that plus the growth on the hardware side, which always comes with a really high margin profile on the software side, that's really what can bring us towards profitability already in the next year. We're looking right now into how can we leverage all the growth opportunities in addition, right, additional markets, additional growth in the defense sector. We're not necessarily optimizing just for profitability. We're optimizing for really taking the big opportunities and winning a global market.
Great, just kind of drilling down on that a little bit and maybe talking about unit economics. I think you previously disclosed each robot, just sort of at this point adds around EUR 100,000 annual recurring software on top of the hardware sale, which I think is around EUR 250,000 kind of per CA-1. I usually try to understand the long-term full picture, like how kind of you view blended margin down the road. How does that improve as you scale production and grow the install base, maybe what the mix might look like, for lack of a better word, steady state, but just at least directionally over the long term, software versus hardware?
Yeah. I mean, at the core of the business model, we are a software business. The hardware right now, specifically for the first years of really rapid rollouts, is the dominating part because it's all about really growing our installed base. We're not really optimizing a lot on the margin profile for the hardware, but more on really decreasing the sales price for the hardware. We really want to make it a commodity. We are winning from really having a massive installed base globally of our autonomous systems, where it's then about collecting, A, the data and, B, the revenues that are coming in from every installed unit, what you mentioned on the EUR 100,000 per year per system. The more systems, the better. The margin profile on the software is really much like 80%+.
That's really much what we're optimizing for, with also more and more modules that we're building, more software solutions along the supply chain, more integrations and more really recurring elements when it comes, for example, to service maintenance support, but also when it comes to licensing different recipes, for example, from famous chefs, from influencers, right? We're right now in the process of adding an entire app store layer, really much for recipes. Plus also in the long term, additional revenue streams through the ingredients themselves, which we are not controlling entirely today, but in the long term is something that we definitely aim for.
It's all about adding more and more revenue streams and kind of upselling the customer through a core of the system that we already have installed at the base. That's pretty much where we're winning from, not the hardware purely itself.
That's helpful. Just to kind of wrap us up here. Just in our last discussion, you laid out a pretty large kind of long-term opportunity for autonomous food systems broadly. It's a big food service market. You've pretty much just focused on the next, say 12 months. What are the two or three milestones in 2026 that would best demonstrate that the model is inflecting?
Yeah, I think when you look into the market today, the most important KPI to look at, I think also from an investor's perspective, is not the amount of units we deploy in a certain amount of time, but it's at what level of quality we're deploying the units, right? What kind of customers, which is all tier one companies, some of the largest enterprises in the world, from Meta to REWE to Mercedes-Benz, German Armed Forces, just to name a couple we talked about. That's really something that is, I think, very important to look at, Nick, when it comes then to the scalability of the accounts. Because right now what is the big chunk of work from the start is the integration of these customers. Scaling to the next location is then an easy part once you've integrated.
It's really about the quality of the customers, obviously metrics like uptime, ecosystem stability, and so on. That's very much what we're improving on. It's also where we share a ton of more insights in our next quarterly update call, which is upcoming in right now two weeks roughly, where we really can share how great the unit is performing. We had a great uptime, far beyond 90% of uptime incidents, which really cut down right now over the entire last year radically. The machine and system-related bugs that you always have in the very early phase of the rollout are right now really at a fraction. That's something that I think is milestones and KPIs that we're tracking, plus the overall growth when it comes to new sectors. We're right now entering the sector and the market of care, defense right now growing.
There's a ton of milestones there on the growth and deployment into new markets and new customers.
All right, very exciting stuff, I'll be, listening intently to your update for Q1. I think it's April 16th.
Yeah.
Thank you again, Nikolas. We really appreciate you joining us for this session of the WTR Insights Conference.
Yeah.
Yeah. Thank you to everyone who participated and dialed in as well. Look for additional content on Circus SE at www.watertowerresearch.com. For those with early questions or for investors wishing to meet with management after this event, please reflect that interest again through the conference portal. Our next conference session will be starting shortly, and we invite you to stay with us. Thank you.
Thank you.