Personal Statement
I am a Lecturer in Physics at the University of Liverpool and a member of the Cockcroft Institute, where I develop machine learning methods for the optimisation and autonomous control of complex scientific systems.

Modern scientific facilities are becoming increasingly complex to operate, requiring experts to navigate high-dimensional control spaces, time-consuming optimisation procedures, and ever-growing volumes of data. While these challenges can often be addressed through manual tuning and conventional algorithms, doing so demands substantial expertise and limits the speed at which scientific discoveries can be made. My research aims to develop reliable, sample-efficient, and trustworthy artificial intelligence methods that augment human expertise by automating repetitive optimisation and control tasks. By enabling scientists to focus on scientific questions rather than routine operation, these technologies have the potential to significantly increase the productivity and capabilities of next-generation research facilities.

I am particularly interested in the application of reinforcement learning to real world systems (e.g., particle accelerators), a unique paradigm inspired by fundamental principles of animal and human behaviour. This interest led me to found the Reinforcement Learning for Autonomous Accelerators (RL4AA) international collaboration, which organises targeted workshops featuring state-of-the-art tutorials each year. The RL4AA'26 workshop was held in Liverpool, with a great coding challenge on RL deployment in CLARA, the linear particle accelerator at the Cockcroft Institute.

I am also interested in Bayesian optimisation and differentiable simulations.

I welcome collaborations across machine learning and experimental science, particularly where reinforcement learning can enable autonomous or decision-support capabilities in real-world systems.

I sporadically post on social media, you can find me on instagram.