Postdoctoral Research Associate in AI for Fundamental Physics and Cosmology, University of Cambridge (Handley Lab)

Job Description

The University of Cambridge is recruiting a postdoc in AI for fundamental physics in the Handley Lab at the Kavli Institute for Cosmology, Institute of Astronomy, for a research post of about four years (reference LG51052).

At a glance

• Position: Research Associate (Grade 7) or Assistant Research Professor (Grade 9)
• Where: Kavli Institute for Cosmology, Institute of Astronomy, Cambridge, UK
• Pay: £37,694–£59,966 a year, depending on grade
• Duration: About four years; funds available until 30 September 2030
• Apply: 25 September 2026 (23:59 GMT, as stated in the ad)

About the AI for fundamental physics project

• Derive physical predictions from new theories of gravitation, fields or cosmology and follow them through to observable quantities
• Direct the building of the research software needed to test those predictions, using AI-assisted development and GPU computing
• Compare the resulting models against cosmological and astrophysical data
• Develop GPU-accelerated Bayesian inference methods, including nested sampling
• Publish and present the research, and take part in the fundamental physics, cosmology and astronomy communities in Cambridge

The ad makes a pointed argument: testing a new theory in cosmology is so much harder than re-running an established model that the field tends to study the theories that are easiest to code, not the most promising ones. The project aims to change that by letting a small group use AI agents and GPU inference to carry a new idea all the way from a Lagrangian to a comparison with data. The scientific judgement stays with the researcher, and the post carries more research freedom than usual: candidates may shape the programme around their own interests.

Why this role matters

Most “AI for science” postdocs ask physicists to become machine-learning engineers. This one asks the opposite: it values theoretical depth first, and treats AI agents as tools a good theorist directs and checks. For a cosmologist or field theorist who wants to test ideas outside the big collaborations, it is an unusual chance to work that way from the start.

What University of Cambridge is looking for

• A PhD in physics, astronomy, applied mathematics or a closely related field, completed or nearly completed
• Strong theoretical ability, which the ad calls the essential criterion
• Background in any relevant area, for example gravitation, field theory, lattice or numerical field theory, cosmological perturbation theory, Bayesian computation, or GPU and differentiable programming
• The judgement to direct AI coding agents, check their output and correct the implementation; the ad rates this above raw programming strength

Pay and location

The salary range is £37,694 to £59,966 a year. Appointment is at Grade 7 (Research Associate) or Grade 9 (Assistant Research Professor) depending on skills, experience and research record; the ad notes that Grade 9 needs Faculty Board approval, and if that is not granted the appointment is made at Grade 7. The post is fixed term, with funds available until 30 September 2030. The University notes that it must make sure every employee is eligible to live and work in the UK, so applicants from outside the UK should read its guidance on visas before accepting an offer.

How to apply

Apply through the official University of Cambridge page. The closing date is 25 September 2026 (23:59 GMT, as stated in the ad). Apply online with a cover letter, CV, publication list, research statement, a GitHub username (or other evidence of computational work) and two academic referees, one of them your most recent line manager. Quote reference LG51052.

See all our postdoc positions, or browse more research jobs on ResearchJobs.in.