Since 2023 Leonhard Liu is part of the Data Science group at BHT, supervised by Prof. Dr. Patrick Erdelt and affiliated with the Hasso Plattner Institute at the University of Potsdam.

His work focuses on machine learning, agentic systems, and the integration of AI workflows with database technologies. In both teaching and research, he is particularly interested in how modern LLMs can interact with data infrastructures in practical, reliable ways.

He has contributed to projects such as medBERT.de: A comprehensive German BERT model for the medical domain (Expert Systems with Applications, 2024) and the study Benchmarking Machine Learning Pipelines in PostgreSQL with TPCx-AI, published in the LNCS series (Springer, 2024).

Outside of work, Leonhard enjoys travelling, spending time with friends and family, and playing the guitar - with an old soul and a young researcher’s curiosity.