Summary
Andrew Thomson is a PhD student and data scientist based in London with 13 years’ experience building optimisation and machine learning systems that drive real-world decision-making under uncertainty. He specialises in end-to-end modelling pipelines combining simulation, statistical methods and ML—applying neural networks, reinforcement learning and multi-objective constrained optimisation to explore high-dimensional chemical spaces for supramolecular design. Comfortable from scientific scripting in Python to high-performance C and legacy Fortran adaptation, he has led cloud-based optimisation on Oracle Cloud with tenancy-level budget and policy ownership and HPC/GPU workflow orchestration (SLURM/SGE). He also teaches practical programming and ML as a PGTA and has delivered outreach workshops, reflecting a knack for translating complex computational ideas to diverse audiences. A less obvious strength is his focus on balancing model performance with compute cost and delivery constraints, making his work both scientifically rigorous and operationally pragmatic.
13 years of coding experience
Master of Physics with Honours, Physics, Master of Physics with Honours, Physics at University of Exeter