Current Projects
Abstract:
Within the CAIMed Research Cluster, this initiative develops causal and interpretable AI models for clinical decision-making by integrating statistical relationships with cause-and-effect mechanisms from medical data. The goal is to build models that reveal underlying biological processes and dependencies, reduce data bias, and support transparent clinical reasoning by combining causal approaches with machine learning and knowledge graphs.
I lead the design of trustworthy AI (LLM) systems for medical applications, working closely with clinical and research collaborators across the cluster.
Role: Project Lead (Trustworthy LLM Systems for Medicine)
Abstract:
SOOFI is a collaborative German research initiative developing a powerful, open-source large language model made available to business and society. Bringing together leading research institutions and innovative startups, the project aims to build an independent, trustworthy European AI technology that integrates specialized reasoning to improve performance while reducing resource consumption.
Within SOOFI, I lead the LLM safety alignment work, ensuring the model meets high standards of safety and reliability for real-world deployment.
Role: LLM Safety Alignment Lead
