Hyllus Systems is a European defense-tech startup specializing in strike drones — kamikaze, bomber, and interceptor systems — designed and iterated in close collaboration with frontline users in Ukraine. Operating between the Netherlands and Ukraine, the company is led by Albin Sela.
We sat down with Albin to discuss what it truly means to build an attack drone company today. From rapid product development using battlefield feedback to Europe’s defense-tech gap, this conversation covers it all.
What Does Hyllus Systems Do, and What Problem Are You Solving?
Hyllus Systems is a defense startup now entering the scale-up phase. We focus on the development of drones and drone systems—specifically, strike drones: kamikaze, bombers, and interceptors. That’s already a niche, especially in the Netherlands, where most players focus on reconnaissance and surveillance. We’re among the few building strike drones.
The problem we’re solving is European dependence on American systems and other countries, both at the system level and in components. Modern warfare is shifting from high-ticket assets (tanks, aircrafts) toward ecosystems of smaller assets: drones, detection systems, ground robots, satellite imagery—still supported by high-ticket platforms.
Europe is too dependent on others for defense. Until 2023, roughly 80% of Europe’s defense budget flowed to American companies—painful. We’re allies, sure, but we’re extremely exposed. Public opinion is finally realizing how dependency can make us almost hostages to American systems—e.g., the ability to switch off fighter jets at the push of a button.
Europe is entering a rearmament phase over the next 3–5 years, with close to a trillion euros being deployed—but today’s European industry isn’t structured to absorb orders at that scale. We need new companies. Some have grown a lot—like Quantum System and Tekever —but they aren’t enough.
We need more companies like Hyllus Systems, with a tech base and doctrinal understanding, to serve as a foundation for European tech development.
Our goal is to build an alternative in Europe to American systems. For example, we rely heavily on American air defense systems. In Europe, beyond MBDA, there aren’t many options. And as we just saw in Poland, you can’t spend multiple million euros to shoot down a $20,000 drone.
That’s changing, but it’ll take time. Startups like ours will drive the doctrinal shift.
How Do Doctrine and Proximity to End-Users Shape Your Products?
First, everything must apply to European armies we work with—the Dutch, Belgian, German, and so on. The key is to create products for us, by us, applicable to real missions and real end-users. If you’re getting bombarded by Russians, you don’t want gear that doesn’t make sense to counter that threat. It must feel natural and spontaneous.
The Russians are doing this. They have an institute—Military Innovative Technopolis “ERA”—with hundreds of scientists and engineers doing lessons learned from hybrid and frontline combat. They funnel frontline feedback back into tech and doctrine, adapting quickly.
If you don’t have close proximity to users—in my view, if you’re not in Ukraine, you’re not really developing, because you don’t know what’s happening.
Trends shift fast. Last October, fiber-optic drones seemed to be the thing—everyone used them. Then we saw a shift: fiber-optic drones are still excellent, but now there is more emphasis on AI-driven drones.
Doctrine is changing within a year, with a significant impact on operations. If you’re not there to see it, you’ll be late—everything takes 2–3 months to catch up. Being close to end-users—today in Ukraine, tomorrow somewhere else—is crucial.
Second, supply chain. People say “just buy in China”—not feasible at large numbers, affordable pricing, and quality. China today isn’t the China we knew; it controls a massive share of materials for components—around 89% for certain materials we need for drones.
Even if we build European factories tomorrow, we will still face raw material constraints. Supply chain will be a key success factor over the next 2–5 years, and it’s partly a matter of politics.
Third, collaboration with government. We need to be close to MODs—ministers, generals, soldiers—keeping alignment with strategic direction. Startups can’t afford 6- to 18-month misalignments: we’re already “slow” compared to battlefield tempo.
Public-private proximity is essential—not only for classic military engagements but to protect critical infrastructure: bases, energy, cyber. I’m not saying startups are the solution, but to help, we must be able to talk properly with MODs.
Additionally, a unified European procurement process, with unified procedures and joint orders, would be beneficial. Things move slowly, but maybe we can push them to accelerate.
How Do You Validate Products in the Field? What Are Your Success Metrics?
We work in levels. We start from feedback—partners, tech experience—build systems, then hand them to the army. They tell us what works and what doesn’t. Iteration cycles typically take around 8 weeks, but it’s never-ending: a product is “good” for maybe two months; every two weeks, something changes, even minorly. That’s why we keep the team small and flexible—if it doesn’t make sense on the battlefield, we change it.
Metrics depend on the product. We’re integrating AI modules (navigation, targeting) across platforms, prioritizing interceptors. Metrics include precision rate, mission accomplishment, and also logistics metrics: component availability, product need, and ease of assembly/build.
Many Ukrainian units receive drone kits rather than fully assembled systems. The assembly and service experience matters. There’s a big difference between assembling in a Dutch workshop with AC versus in a trench in Ukraine.
Ease-of-use under battle conditions is critical.
Biggest Business Risks That Keep You up at Night?
Number one: supply chain—huge issue. Number two: working with MODs. Many ministries are reforming and adapting; the Netherlands is creating task forces and startup programs. Great, but most MODs still don’t know how to work with small companies after 30 years of buying from primes.
Tendering is a problem: there is still a significant lack of tenders for drone procurement in European armies. It makes sense—the tech becomes obsolete in six months. But for a small company that can’t sell to private buyers, that’s a risk.
Regulations and compliance, especially EU AI rules. We’re working with partners in Europe and Ukraine; targeting modules have been on the market for around a year, but are still not extremely advanced. We need more data and training.
We already know that you can’t deploy a model that can’t reliably distinguish a person from a dog, or that lacks a proper human-in-the-loop / arming interlock. Some requirements make sense in principle, but we need a way to apply them while staying competitive—Russians already have strong navigation models, and the Chinese are in a different league. We must catch up fast.
Fundraising: VCs vs. Family Offices?
We started preparing for the fundraising process over the summer. I needed something concrete to put on the table—products, traction, readiness. However, there’s a significant cultural difference: U.S. investors understand and accept risk capital, whereas in Europe, many want to be certain they will get their money back tomorrow.
Also, many European VCs were SaaS-only until recently; now they talk hardware/defense using SaaS metrics. Different business. I don’t like the short-term view—I want partners in it for the long run.
So we chose family offices and individual investors. They’re pragmatic and understand the difference between B2B SaaS and defense-tech. We didn’t talk to VCs; since late summer, we’ve been going the family-office route.
It’s not easy—there’s a lot of noise in the drone market, and it doesn’t take much to build a drone. But combat drones are different: what’s your hit rate in Ukraine? Do you have confirmed effects? That separates real systems from hobby builds.
Also, we avoid over-engineering. I’ve heard major VCs say, “It’s so simple,” but it took us 12 months to make it simple. That’s the point: simple, effective, fewer points of failure. If you want scale, it must be affordable, available, and effective. The more complex it is, the worse it performs under stress. SpaceX’s mantra applies: simplify, simplify, simplify.
We’re talking to some funds and family offices—that’s the direction. I don’t think most VCs are ready yet.
What About Recruiting and Retaining Top Technical/Operational Talent?
My “sales line” is rough but true: we’re not here to chase a startup success story. We need to help win the war and prepare Europe for tomorrow—and “tomorrow” is very soon. If you join us, you’re joining a mission. We work like crazy because we believe in it.
Our people could earn three times the money elsewhere, but they stay for the mission. Young engineers want purpose, not just money. They’re tired of corporate. Startups can be great when they attract motivated, smart people who believe in what they’re doing. Give work real meaning, and people go further—because they believe in it.
We also built genuine relationships with Ukrainian units. It started as a business and became a friendship. We want them to win. That motivation is our main retention driver.