I'm a recent PhD graduate from the Transport Systems and Logistics Laboratory at Imperial College London, where I was advised by Washington Ochieng, Mohammed Quddus and Panagiotis Angeloudis.
I'm research interest lie in robot learning and decision making, with a focus on modelling interactions between different agents like autonomous vehicles and other mobile robots. During my PhD, I spent time as a research intern at NVIDIA's Autonomous Vehicle Research Group, where I worked on developing information-theoretic learning algorithms to mitigate the sim-to-real gap when training a policy on synthetic data.
I believe that the road to widespread Physical AI requires scalable systems, from simple models to efficient data and compute utilisation. Simulation and world models offer a powerful path to scaling autonomous systems, allowing robots to perceive, reason, and generalize beyond limited real-world data. My goal is to develop methods that advance this frontier by maximizing the value of available data and compute.
I subscribe to a pragmatic approach to research, often attributed to Richard Feynman: "What I cannot create, I do not understand". On my blog, I document my learning by building and experimenting with ideas, from first principles to applied systems.