I am a DPhil researcher at the University of Oxford, within the Visual Geometry Group (VGG), supervised by Professor Andrea Vedaldi. My research focuses on Spatial AI, 3D computer vision, and controllable 3D reconstruction, particularly learning representations for geometric and physical reasoning.
Previously, I worked as a Machine Learning Engineer at Mercedes-AMG Petronas Formula One Team, researching and developing deep learning methods for aerodynamic modelling, computational geometry, and scientific simulation. My work included transformer architectures for unstructured 3D meshes and point clouds, aerodynamic prediction, inverse design, and GPU-accelerated simulation infrastructure using CUDA, C++, PyTorch, and NVIDIA Warp.
My background also includes software engineering, architecture, visual effects, procedural generation, and computational design. My broader interests include computer graphics, robotics, world models, and high-performance machine learning systems.



