This German AI Startup Just Raised $13.2M to Build AI That Learns Like Humans Do

Jan de Vries · 2026-07-23

### The Big Picture: Why This Matters Most AI today is like a brilliant student who can only answer questions from the textbook they memorized. Change the context even slightly, and they're lost. That's the problem kausable, a German DeepTech startup, is tackling head-on. They just raised $13.2 million in a Seed funding round to build a reasoning-first AI that adapts without needing constant retraining. ### The Funding Details The round was led by UVC Partners and Entourage, with follow-on investments from German firms HTGF and Mätch VC. They're also backed by angel investors from top AI labs like Black Forest Labs, OpenAI, Google DeepMind, Noxtua, and the European Laboratory for Learning and Intelligent Systems (ELLIS). This is serious backing from serious people. ### What Makes kausable Different? Founders Johannes Haux (CEO), Dr. Benjamin Herdeanu (CTO), and Gregor Ramien (COO) are three physicists from Heidelberg University and Black Forest Labs. They're building what they call a "world model" — a foundational AI that learns from minimal data and adapts to situations it couldn't have anticipated during training. "The world evolves rapidly, and so should AI," says Haux. "Current systems need constant, costly retraining. We're solving that." ### The Core Problem Today's AI systems ingest massive datasets and retrain frequently, hoping the world doesn't shift too much in between. kausable's approach is different: their model adapts as conditions change. It can merge different data sources into a single comprehension layer and operate in scenarios that weren't foreseeable at build time. ### Real-World Applications This isn't just academic. The potential is huge across dynamic fields: - **Robotics**: Robots that adapt to new environments without reprogramming - **Energy sector**: Predicting grid fluctuations and optimizing in real-time - **Finance**: Spotting market shifts before they become trends They've already developed TipPFN, a forecasting zero-shot model that predicts "black swan" events — rare but highly impactful occurrences in areas like medicine and energy. ### The Human Touch kausable's reasoning-first AI learns like humans do. Think of it as having a set of universal intuitions that help it adapt quickly to new situations with minimal new information. The team recently co-authored a research paper with experts from Columbia University, validating the causal reasoning architecture behind their model. "Most AI models are trained to remember the past. kausable is building AI that can reason about the future," says Pieterjan Bouten, co-founder of Entourage. "It's an ambitious scientific bet." ### What's Next? The company plans to use this funding to expand its current nine-person team and advance its rapid-learning frontier model. They've also raised a pre-Seed round of $1.65 million, showing strong early investor confidence. ### The Bottom Line In a world where AI is often about bigger models and more data, kausable is taking a different path. They're betting that reasoning and adaptability will win over brute force. And with this funding, they have the resources to prove it. *This article was originally featured on EU-Startups.*