First Light Fusion Ltd is a pioneering UK-based fusion energy company at the forefront of developing breakthrough inertial fusion technology. At the heart of our mission is FLARE — First Light’s next-generation fusion approach designed to accelerate the path toward practical, clean, and limitless energy. By combining world-leading expertise in advanced physics, simulation & machine learning, and innovative engineering, FLARE represents a bold step forward in delivering scalable fusion solutions.

Alongside our fusion program, First Light Fusion also delivers commercial services and partnerships applying our simulation tools and hydrodynamic amplification technology to solve complex challenges across a range of advanced industries.

Joining First Light Fusion means being part of an ambitious team shaping one of the most exciting energy transformations of our time.

Job Description

We are looking for a skilled and enthusiastic scientist to join our Data-Driven Engineering Team, who is interested in applying data science and machine learning to physics and engineering problems. The position is available at the scientist level, but we may consider exceptional candidates at the senior scientist level.

You will have practical knowledge of the application of statistics and machine learning to scientific data analysis and simulation-based modelling. You will collaborate with a highly skilled team of physicists and engineers to develop and implement ML tools that will inform critical decisions about our fusion targets and hydrodynamic amplifier technology. First Light Fusion operates its own experimental and manufacturing facilities, enabling us to move quickly and innovate. This will allow you to test your models quickly against real data within the design cycle, enabling time-critical decisions to be made. Additionally, you will support the ongoing development of our data science software stack by researching and implementing leading-edge ideas in machine learning and applied statistics. You will have the opportunity to communicate your work at conferences and in publications.

Accountabilities and deliverables:

• Research and implement machine learning technologies that further develop our capability to design fusion targets and hydrodynamic amplifier technology.
• Close the design loop by working across computational physics, design engineering and experimental physics.
• Contribute to the ongoing development of our data science software stack, for which you will also act as a point of contact in the company.
• Communicate and record your work in technical meetings and written reports.

Core skills, knowledge and attributes:

Essential

• PhD in a STEM subject, or equivalent industry experience
• Successful track record of planning and executing research in data science and machine learning
• Hands on programming experience using Python and libraries that are relevant to data science and machine leaning (such as Pandas, PyTorch & TensorFlow)
• Experience working with High Performance Computing systems
• Strong communication and interpersonal skills
• Ability to collaborate effectively with a diverse team of scientists and engineers

Desirable

• Knowledge of physics, particularly in high energy density plasma physics and inertial confinement fusion.
• Experience developing software for scientific data analysis and modelling
• Experience using hydrodynamics simulation and modelling
• Experience working with at least one of the following:

ML surrogates, including uncertainty-aware models, such as Gaussian Processes

Optimization methods such as evolutionary strategies and Bayesian optimization

Bayesian inference and its application to model calibration and uncertainty quantification

Package

• Competitive salary
• 25 days annual leave + bank holidays (holiday increase based on service)
• 40 hour working week
• Office based in Oxford (flexible working options may be available on request)
• Free lunch, snacks and soft drinks
• Free parking onsite
• Cycle to work scheme
• Electric vehicle scheme
• 8% employer pension contribution without matching requirements