SEREACT raises €93 million to position itself at the core of real-world AI

By raising €93 million in a Series B round, Sereact is positioning itself within an accelerating phase of AI applied to the physical world, where the key challenge is no longer limited to robot performance, but extends to the standardization of their intelligence.

Founded in Stuttgart, the company is developing a distinctive approach: decoupling the robot’s brain from its body. While the industry has long favored tightly integrated systems, Sereact introduces a software layer capable of running across different machines, regardless of their hardware architecture.

From picking to planning: an evolution in the role of AI

Sereact’s core product, Cortex, is designed as an embedded intelligence system for robots operating in real-world environments. Its latest version, Cortex 2, introduces a notable shift: the integration of a world model that allows the system to anticipate the consequences of an action before executing it.

This approach brings robotics closer to the dynamics observed in large language models, by introducing capabilities for simulation and reasoning. AI no longer simply reacts to a situation; it evaluates multiple possible scenarios and selects the one with the highest probability of success.

This capability opens up use cases beyond traditional warehouse picking. Handling fragile objects, constrained assembly tasks, or operations requiring precise orientation are all areas where physical interaction and environmental uncertainty demand a more refined understanding of the real world.

The warehouse as a learning environment

The initial focus on logistics warehouses provides a particularly data-rich environment, combining object diversity, strong operational constraints, and immediate consequences in case of error.

Sereact reports more than 200 systems deployed in production across Europe, totaling over one billion picking operations. This volume underpins a continuous learning loop, where each interaction contributes to improving the model.

A European dynamic gaining momentum

Sereact’s funding comes amid intensifying investment across Europe in physical AI and robotics. Several players have recently raised capital across complementary segments, including warehouse automation, robotic skill platforms, data infrastructure, and physical simulation.

Within this landscape, Sereact stands out through its cross-functional positioning focused on the software layer. Rather than developing proprietary robots, the company follows an interoperability-driven approach, aiming to equip an existing base of industrial machines.

Toward hardware commoditization?

Sereact’s thesis points to a gradual shift in the value chain. As robotic components become increasingly standardized, differentiation is expected to move toward software and the ability to leverage real-world data.

In this context, performance no longer depends solely on mechanical quality, but on the model’s ability to generalize, learn, and adapt across varied environments.

International expansion in sight

Beyond the European market, Sereact is beginning its expansion into North America, with plans to open an office in Boston. This move is intended to capture growing demand for automation solutions in logistics and industrial supply chains, particularly in a context of labor shortages and increasing pressure on operational efficiency.

The company also reports rising interest in the United States following recent demonstrations of its technology at industry events.

Funding to structure the scale-up phase

Sereact has raised €93 million in a Series B round led by Headline, with participation from Bullhound Capital, Daphni, and Felix Capital, alongside existing investors Air Street Capital, Creandum, and Point Nine. This follows a €25 million Series A completed in 2025.

Founded in 2021 in Stuttgart by Ralf Gulde and Marc Tuscher, the company is developing an artificial intelligence platform for robots operating in physical environments, with the objective of delivering a unified model capable of running across different types of machines. The funds will support the development of Cortex 2, team expansion, and international growth, particularly in the United States.