Sereact raises $110M for the intelligence layer behind warehouse robots

Sereact, a Stuttgart-based robotics software company, has raised $110 million in Series B funding to scale its robot intelligence platform and expand into the U.S.

The round was led by Headline, with participation from Bullhound Capital, Daphni and Felix Capital. Existing investors Air Street Capital, Creandum and Point Nine also returned. Sereact says the funding will support Cortex 2.0, its latest robotics AI model, along with U.S. expansion.

In terms of warehouse robotics, Sereact sits in the stack at the software layer creating systems that can run across robotic systems and make manipulation less brittle in real operating environments.

Warehouse automation has spent years moving from fixed, repetitive machinery toward more flexible robotic picking and handling. The hard part is getting the system to handle variation, recover from errors, work across different items and keep intervention rates low enough for production use.

Sereact’s original wedge was warehouse and manufacturing automation, especially pick-and-pack work. The company was founded in 2021 by Ralf Gulde and Marc Tuscher and raised a $5 million seed round in 2023, followed by a €25 million Series A in 2025. That sequence makes the new Series B a fast step up from applied warehouse software into a broader robotics intelligence company.

Cortex 2.0 is being positioned as the next version of that stack. Reporting around the round describes it as a model that helps robots simulate possible outcomes before acting, choosing the physical behavior most likely to succeed. In practical terms, that pushes Sereact toward the same larger question facing the whole sector: how much of robotic manipulation can be generalized across bodies, tasks and environments without rebuilding the system every time.

The company already points to logistics and manufacturing customers, with BMW and Daimler Truck appearing in reporting around its deployments. Sereact’s credibility comes from connecting model development to production environments where error recovery, throughput and reliability decide whether the software is actually useful.