
Ex-F1 Engineer's Robotics Startup Microagi Raises $55M in Germany's Largest-Ever Seed Round
Ten months after founding, the Munich startup led by former Red Bull Racing engineer Bercan Kilic has landed a record seed led by Hummingbird to train factory robots on real-world edge cases.
Munich-based Microagi has raised $55 million in what is being called the largest seed round in German startup history — roughly ten months after the company was founded by former Formula 1 engineers.
The Round
Venture firm Hummingbird led the round, with participation from Northzone, LocalGlobe, Village Global and Redalpine. The capital will scale Microagi's data platform across global manufacturing conglomerates, fund the expansion of its newly opened research headquarters in Zurich, and accelerate technical hiring.
From Pit Lane to Production Line
Founder and CEO Bercan Kilic spent his previous career as an aerodynamicist at Red Bull Racing before founding Microagi with CTO Nico Nussbaum. The company builds Atlas (no relation to Boston Dynamics' humanoid), a hardware- and model-agnostic software platform that captures real-time production edge cases, expands them in simulation, and fine-tunes robot models for plant-specific tasks — the unglamorous work of moving factory robots from demos into production.
The data engine has an unusual twist: Microagi pays some 20,000 people across 15 countries to film themselves doing everyday work — factory tasks and household chores — building a proprietary corpus of humans manipulating the physical world that robot foundation models desperately lack.
Why Investors Bit
Embodied AI's bottleneck has shifted from models to data and deployment. Humanoid makers from Boston Dynamics to China's Zeroth can build the hardware; what they can't easily get is the plant-specific training data that makes a robot useful on this line, in this factory. Microagi is betting the picks-and-shovels layer — data capture, simulation expansion, safe fine-tuning — becomes the toll road every deployment crosses.
A record German seed for a ten-month-old company says the market agrees, at least for now. The test is whether manufacturing conglomerates — notoriously slow buyers — adopt at venture speed.
Newsletter
Get Lanceum in your inbox
Weekly insights on AI and technology in Asia.


