Sereact raises €93M to build robots that predict their own mistakes
What's the deal? German robotics software startup Sereact has raised €93 million ($110 million) in a Series B round led by Headline, with Bullhound CapitalDealroom has a profile for this one. Try Dealroom →, Felix CapitalDealroom has a profile for this one. Try Dealroom →, and Daphni among new investors. The Stuttgart-based company develops AI software that lets industrial robots perform tasks they haven't been explicitly trained on — and already powers machines for BMWDealroom has a profile for this one. Try Dealroom → and Daimler TruckDealroom has a profile for this one. Try Dealroom →.
Why now? Most of the funding will accelerate development of Cortex 2.0, Sereact's newest AI model, which enables robots to simulate outcomes and anticipate errors before they occur — much like reasoning models from OpenAI andAnthropicDealroom has a profile for this one. Try Dealroom →, but for the physical world. The company will also expand in the US. Global robotics investment more than tripled year on year to $27.6 billion in 2025, and the sector is drawing serious capital: peer Neura Robotics is reportedly raising about €1 billion.
What could go wrong? Robots remain largely reactive — errors like dropping items are hard to undo, and real-world deployment at scale is still nascent. Cortex 2.0's ability to anticipate such failures hasn't yet been proven at scale.
The signal: Sereact's raise reflects a broader conviction that physical AI is the next frontier. The startup is targeting a specific, high-volume use case — processing returned goods for retailers — with H&M reportedly in talks. As AI models grow smarter and hardware cheaper, the race to make industrial robots genuinely adaptable is intensifying, and software-first players like Sereact are positioning themselves at the centre of it.
Sources:
Bloomberg
The Business Times
Sifted
EU-Startups
A.M.
