European AI Startup Kausable Raises €12 Million in Seed Round to Build AI that Needs No Retraining

Kausable, putting science at the core of AI, led by Johannes Haux, Dr. Benjamin Herdeanu, and Gregor Ramien, has raised €12 million in a seed funding round led by UVC Partners and Entourage, with follow-on from the German investors HTGF and Mätch VC. Kausable is also backed by various private angel investors from the AI industry and academia working at Black Forest Labs, OpenAI, Google DeepMind, Noxtua, and the European Laboratory for Learning and Intelligent Systems (ELLIS).

Kausable is developing reasoning-first frontier AI that can adapt to changing contexts without the costly retraining required by today's AI systems. The company, which has ties to Heidelberg University and Black Forest Labs, aims to build AI models that continuously generalize and reason in dynamic environments.

The funding comes amid growing demand for digital sovereignty, as geopolitical tensions increase the need for frontier AI models and critical AI infrastructure developed and controlled within Europe.

kausable is developing a new generation of reasoning-first AI that can continuously adapt to changing environments without constant retraining, enabling more efficient AI deployment across dynamic industries such as robotics, energy, and finance. The company believes the approach can transform AI from expensive, one-off implementations into scalable systems for industrial applications.

Backed by European investors including UVC Partners, Entourage, HTGF, and Mätch VC, the company also sees its funding as a signal that Europe can build and support globally competitive frontier AI models. Investors believe kausable's focus on reasoning, causal inference, and self-adapting AI represents a fundamental shift beyond today's data-intensive AI systems.

Founded in 2025 by Johannes Haux, Dr. Benjamin Herdeanu, and Gregor Ramien, kausable emerged from research at Heidelberg University and experience in startups, cybersecurity, and banking. After raising a €1.5 million pre-seed round, the company developed TipPFN, a zero-shot forecasting model designed to predict rare, high-impact events across sectors such as healthcare and energy.

kausable's reasoning-first AI is built around a "world model" that enables rapid adaptation to new environments with minimal additional data, drawing inspiration from how humans learn. The company has also collaborated with researchers from Columbia University to validate the causal reasoning architecture behind its models. The new funding will support team expansion and accelerate development of its frontier AI platform.