Piotr is the head of AI at RealmLabs where he is leveraging research and practice in machine
learning interpretability and security for protecting data in the face of LLM apps tuned to
overshare. Previously he worked at Truera where he developed TruLens, an open-source toolkit for
evaluating LLM applications. Prior to Truera, he was a researcher at Carnegie Mellon University and
University of Maryland where his research included interpretability and accountability for privacy
and fairness in machine learning pipelines; and probabilistic programming and formal methods.