Alex Sopio PhD
Researcher in High Energy Physics & AI
University of Edinburgh, Scotland
About
I am a physicist and AI scientist with a background in high-energy particle physics and a passion for applying machine learning to complex, data-rich problems. My research spans jet physics at the ATLAS experiment, graph neural networks, and quantitative finance.
I enjoy building tools at the intersection of physics intuition and modern deep-learning architectures — from LundNet jet taggers to GNNs in Julia. When not crunching collider data, I explore algorithmic trading strategies and write about things I find interesting.
Recent Posts
Highlights
Developed a GNN-based jet tagger using the Lund plane representation, implemented in both PyTorch and Julia's Flux.jl.
Designed and implemented automated trading strategies for the IMC Prosperity challenge, finishing in the top 4%.
Graph Neural Networks in Julia using Flux.jl, with jet construction via Python bindings to FastJet.