Bio
Stefan is a PhD candidate at the University of Cambridge. His research focuses on machine unlearning (i.e., how to make AI forget specific data) and AI/LLM agents. He previously co-founded a European Space Agency Business Incubator (ESA BIC) funded deep tech startup and worked at the Boston Consulting Group. His PhD is funded by the UKRI EPSRC DTP.
Research interests
Stefan’s research blends both theoretical and applied aspects of machine learning. He is particularly interested in foundational questions such as “how do models memorize information?”, which can then be used for real-life applications like the removal of unwanted data from already trained models via machine unlearning. His interests further include privacy and security-related research that is necessary to enable safe and reliable AI/LLM agents.