Job Description
Be Part of Research That Shapes the Future of AI
The School of Informatics at the University of Edinburgh is seeking a Post-Doctoral Research Associate to contribute to ambitious research focused on benchmarking next-generation foundation models.
This role sits within a £2M programme funded by the Advanced Research and Invention Agency (ARIA), under its Scaling Compute: AI at 1/1000th the Cost initiative. The project aims to evaluate and redesign how we measure AI performance — balancing accuracy, cost efficiency, speed, and environmental impact.
You will work under the supervision of Dr Edoardo Ponti and Dr Luo Mai, collaborating with colleagues across Edinburgh, Imperial, and Edinburgh Parallel Computing Centre (EPCC), one of the UK’s leading high-performance computing centres.
What You’ll Work On
- This position combines research innovation with practical systems benchmarking.
Your responsibilities will include:
- Designing and evaluating efficient LLM and VLM architectures
- Developing benchmarks for sparse, reasoning, and agentic foundation models
- Measuring performance trade-offs across hardware platforms
- Analysing cost, energy usage, and scalability metrics
- Collaborating across institutions to support project milestones
- Contributing to publications and international conference presentations
- You will have access to advanced HPC infrastructure and funding for international conference travel.
About the Research Environment
The School of Informatics is recognised as one of Europe’s leading centres for AI, machine learning and NLP research. You’ll join a vibrant academic community with strong interdisciplinary connections and global partnerships.
Flexible working arrangements (hybrid model) may be considered on a non-contractual basis.
About You
We welcome applicants who demonstrate intellectual curiosity, technical strength and collaborative spirit.
Essential background:
- PhD (completed or near completion) in NLP, ML, ML Systems, Computer Science, Engineering, Mathematics, or related field
- Evidence of research excellence (e.g. publications, preprints)
- Experience implementing and deploying foundation models (LLMs/VLMs)
- Familiarity with performance benchmarking and hardware cost tracking
- Ability to work effectively within a multidisciplinary team
- Selection will focus on the overall strength and coherence of your research profile rather than a strict checklist.
What We Offer
Professional & Research Benefits
- Competitive academic salary (UE07 scale)
- Access to high-performance computing facilities
- International travel funding for conferences
- A collaborative, globally respected research environment
Employment Benefits
- Generous annual leave entitlement
- Defined benefit pension scheme
- Staff discounts and wellbeing initiatives
- Flexible and family-friendly policies
- Employee discount schemes
- Company pension
Pension guidance:
Workplace and public service pensions
Holiday entitlement overview:
Statutory annual leave guidance
Equality, Diversity & Inclusion
The University holds a Silver Athena SWAN Award and is committed to advancing equality across gender, race and LGBTQ+ inclusion. We welcome applications from candidates of all backgrounds and are committed to providing an inclusive recruitment process.
Equality legislation overview:
Visa & Right to Work Information
This role is open to UK and international applicants. Visa sponsorship may be available depending on eligibility and individual circumstances.
How to Apply
Please submit:
- Curriculum Vitae
- Cover letter outlining suitability and research interests
- Up to three academic papers demonstrating research quality and alignment with the project
Closing date: 5 March 2026
(Applications close at 11:59pm UK time.)
Why Join Edinburgh?
A role at the University of Edinburgh means joining a diverse, international academic community committed to research excellence and global impact. You will contribute to meaningful innovation in AI systems while developing your academic career within one of the UK’s most respected research institutions.