What kausable does
kausable develops causal AI systems, sometimes called world models, that the company says can adapt to new situations from minimal examples rather than requiring retraining on large datasets. Its first demonstrated model, TipPFN, is a zero-shot forecasting model aimed at predicting rare, high-consequence events in complex dynamic systems. The company is working with early customers in robotics, healthcare, energy and industrial systems. kausable grew out of research at Heidelberg University and was founded in 2025 by Johannes Haux, Dr Benjamin Herdeanu and Gregor Ramien.
kausable is an ai models tool on Falcoscan. Causal 'world models' that adapt to new situations without retraining. Falcoscan rates kausable with an Opportunity score of 79/100, a Saturation score of 30/100, and a Wrapper-risk score of 8/100. Market signal: hot. kausable is founded in 2025, currently at Seed stage. Pricing: Paid.