Email: p.romov25@imperial.ac.uk

Peter joined the AI Security and Privacy Lab in November 2025 as a PhD student. His research interests include privacy-enhancing technologies and the security of LLM-based agents.

Before starting his PhD, Peter spent more than ten years working as an applied machine learning engineer at major technology companies, including Meta, WhatsApp, and Yandex. Peter has also taught machine learning at universities, organised machine learning competitions, and built a community of data science professionals.

At Meta AI, he worked in the privacy research group led by Ilya Mironov, where he contributed to the development and deployment of privacy-preserving technologies across Meta products, including differentially private federated analytics, secure multi-party computation, and differentially private ML.