Research
My work spans high-energy particle physics, graph neural networks, and applications of machine learning to scientific data.
Interests
Studying the internal structure of hadronic jets produced in LHC collisions. Using the Lund plane to represent jet histories for machine-learning taggers.
Applying GNNs to particle-physics datasets, where collision events have a natural graph structure — particles as nodes, relationships as edges.
Designing and backtesting algorithmic trading strategies, with particular interest in market microstructure and statistical arbitrage.
Experience
Developing graph neural network jet taggers for the ATLAS detector at the LHC. Focus on the Lund-plane representation and Julia-based implementations.
Ranked 37th out of more than 1,000 participants by designing and optimising automated trading algorithms for various simulated financial instruments.