What do you do at DYNAMIC?

I am a postdoctoral researcher at the UKP Lab at Technische Universität Darmstadt, where I develop and evaluate AI methods, particularly large language models (LLMs), to automatically construct person-specific psychological networks from therapy transcripts and support culturally sensitive, responsible mental-health applications.

How would you describe your work and research? 

My work develops responsible, culturally sensitive AI for mental health by using large language models to extract personalized psychological networks from therapy sessions, generate privacy-preserving synthetic counseling dialogues, and rigorously evaluate these systems for safe and effective use.

What insights from your field of research have you always wanted to share?

AI has strong potential to expand access to mental health support, but its most meaningful role is to assist clinicians rather than replace them. To ensure these systems are both safe and useful, close collaboration between AI researchers and mental health experts is essential so models are clinically meaningful, ethically grounded, and not just technically sound.