Microagi’s Bercan Kilic raises $55m seed to teach factory robots with chore footage

Bercan Kilic has a pretty specific flex: he left designing F1 aerodynamics for Red Bull Racing while the team was “winning everything,” and turned that fascination with engineering into a new mission. In 2023, he got his dream job, designing aerodynamics for Red Bull Racing. He also noticed something else, which he found “magnificent” but somehow “thin” in the point of it. That tension becomes the through-line for microagi, a Munich startup Kilic leads, now raising $55m.

microagi says it just secured what it calls the largest seed round a German company has achieved. Hummingbird led the round, with Northzone also participating. So this is not just another robots-on-a-slide deck story. It is a very real capital validation moment for a founder who came from elite, performance-obsessed engineering, and now wants factory robots to learn from something more ordinary than race car airflow.

The technical idea described by The Next Web is that microagi is raising money to “teach factory robots.” The company’s unusual ingredient, according to the headline prompt, is footage of people doing chores. That matters because manufacturing robotics has historically been expensive to scale. You do not just need an algorithm. You need data that reflects the world robots will actually touch: irregular objects, imperfect grasps, messy environments, and human movement that is not standardized for machine learning. If chore footage can help bridge that gap, it suggests a practical path: use low-friction, real-world demonstrations to reduce how much bespoke data collection and trial-and-error a factory has to fund.

For executives, there is also an incentive story hiding behind the engineering one. The factory automation market lives at the intersection of capital expenditure and operational pain. Robots are a bet that throughput and consistency will outperform the cost of deployment. But when deployments struggle, the “ROI clock” starts ticking against you: downtime, integration cost, training cost, and change management all compound. A training approach that leans on more readily available footage could lower the total cost of readiness. Even if the headline only hints at the method, the board-level implication is straightforward: any reduction in onboarding friction is a competitive advantage, because it can make pilots easier to justify and scale faster if results show up.

There is a second layer too: the pedigree-to-capital translation. Kilic spent time at Red Bull Racing designing aerodynamics, during a period when the team was dominant. Motorsport is, in a sense, a laboratory where engineers are rewarded for compressing the gap between theory and real-world performance. When a founder with that background walks into a seed round, investors are not just buying an abstract thesis about robotics. They are betting that the founder understands how to iterate quickly, measure outcomes, and ship systems that survive contact with reality. microagi’s fundraising, led by Hummingbird with Northzone involved, reads like the market agreeing that engineering credibility can be a lever, especially in early-stage “hard tech” where execution risk is the default.

Now add the broader context that typically shapes robotics capital decisions. In most ecosystems, robotics and AI get funded heavily when the industry sees a clear path to scale. Early-stage funding can be won by teams that can articulate how learning improves deployment speed and reduces uncertainty. If chore footage is a stand-in for real operational variation, it is a way of arguing that the learning loop is not fantasy. It is anchored in the messy distribution of human activity that factories already contain, even if that footage has not always been treated as training material.

There are also governance and oversight considerations that boards and CFOs tend to care about when robots move into physical spaces. While the source does not spell out regulatory steps, the underlying reality is that factory deployments require safety planning, risk assessment, and controls around how systems behave when they encounter unexpected situations. A training approach that expands the robot’s experience before it ever leaves the lab could, in theory, reduce the number of hazardous edge cases discovered late. That is not a guarantee, but it is exactly the type of “preparation advantage” that makes investors comfortable in a seed round: it signals thoughtfulness about how you get from prototype to production without burning schedule and trust.

The strategic stakes for peers in similar roles are immediate. microagi is claiming a high bar for German seed funding, and it is doing it with a founder profile tied to F1 engineering and a method tied to everyday footage. That combination creates a signal for other builders: the next wave of automation may not only be about more compute or prettier models, it may be about smarter training data strategies that make robots easier to deploy. If microagi can convert $55m into working factory outcomes, it could force competitors to rethink how they source data, how they reduce pilot risk, and how they explain the path from demos to scaled systems.