About Me
Welcome! I am a PhD Candidate in Artificial Intelligence at Imperial College London. My research is focused on improving robustness and generalisation in deep reinforcement learning. I am currently working on two questions:
- Policy-aware world models for AI safety evaluation: Can we efficiently use knowledge about a deployment policy to learn a world model that would accurately predict safety of an agent’s behaviour?
- Safe neural policy updates in continual learning: Can we update a neural policy to a new task while preserving its safety/robustness guarantees in prior tasks?
Besides, I am interested in ML and stats applications in quantitative finance. Currently, I am doing a PhD internship at Cubist Systematic Strategies (Point72) in London working on ML research for systematic global macro.
At Imperial, I am one of the organisers of the Imperial College Autonomous Reasoning & Learning (ICARL) group, helping manage research seminars and reading groups. I am also a part of the Centre for Doctoral Training in Safe & Trusted AI.
Prior to my doctorate studies, I spent 2.5 years as a data scientist at causaLens and worked as a quant in asset management in London and the Netherlands for 1.5 years. I earned my MSc in Econometrics at Erasmus University Rotterdam and BSc (Hons) at Higher School of Economics, majoring in Mathematical Economics and minoring in Data Science & Machine Learning.
I hold a UK Global Talent visa as an emerging leader in AI & ML. I have received multiple academic merit-based scholarships, including full funding for my PhD and the Holland Scholarship for excellent incoming master’s students from outside of the EEA.