Applications are invited for aResearch Software Engineer (SC6) orSenior Research Software Engineer (SC5) to join Professor Ke Li’s newly establishedAI for Biology Group at the Earlham Institute, Norwich, UK. This is a full-time, 36-month position, with the possibility of extension.
The group builds on Professor Li’s established and well-funded programme in fundamental AI, biological foundation models, AI co-scientists and automated scientific discovery. At Earlham, it will develop an ambitiousGenerative Digital Biology programme: experimentally grounded AI systems that learn from biological data, generate hypotheses, guide experimental design and accelerate discovery.
The role
We seek a technically outstanding, research-focused software engineer to design and build the programme’sAI-agent platform. You will connect AI models, scientific tools, biological datasets, experimental-design methods, compute infrastructure, laboratory automation and human decision-making into robust systems for AI-driven biological discovery.
The work may include agent orchestration, tool registries, model and data interfaces, workflow execution, retrieval-augmented generation, versioning and provenance, evaluation, permissioning, sandboxing and safe tool use. You will develop reusable agentic workflows and an extensible platform for scientific reasoning, hypothesis generation, experimental planning, result interpretation and human–AI collaboration.
This is not a conventional software-support role. You will have substantial scope to shape the technical architecture, lead high-quality open-source software, contribute to publications and benchmarks, and help establish a long-term AI co-scientist capability. The role involves close collaboration with theEarlham Biofoundry, Research e-Infrastructure, BioFAIR, ELIXIR-UK and biological teams across the Institute.
Research environment
You will have access to dedicatedNVIDIA H200 GPUs, additional capacity through the Norwich Data Centre, and routes to national-scale AI compute includingIsambard-AI. The Earlham Biofoundry provides opportunities to integrate AI systems with laboratory automation and design-build-test-learn workflows.
About you
You will have a PhD, or equivalent research or industrial experience, in Computer Science, AI, Robotics, Machine Learning, Software Engineering, Scientific Computing, Data Science, Computational Biology or a related discipline.
You should demonstrate strong hands-on software engineering ability and experience building complex research software, AI systems, agentic workflows, scientific platforms, robotics or data-intensive infrastructure. Essential capabilities include robust, reproducible development using version control, testing, documentation, containers, APIs, databases, workflow systems and CI/CD, together with deployment experience on HPC, GPU, cloud or distributed environments.
Applicants should provide evidence of high-quality technical or research outputs appropriate to career stage, such as open-source software, widely used tools, benchmarks, prototypes, demonstrators, patents, technical reports or publications. Experience with large language models, tool-using agents, retrieval-augmented generation, workflow orchestration or model APIs is expected. Biology, robotics or laboratory-automation experience is advantageous but not essential. A commitment to trustworthy and responsible AI is required.
Appointment at SC5 requires additional evidence of technical leadership, end-to-end ownership of complex systems, delivery with limited supervision, leadership of major outputs, coordination of contributors, mentoring, and contribution to platform strategy and grant applications.
Salary: £38,000–£46,500 at SC6 or £47,450–£52,560 at SC5, depending on qualifications and experience.
This role meets the criteria for a visa application. Earlham Institute is a Disability Confident employer and guarantees an interview to disabled applicants who meet the essential criteria.
Closing date: 4 September 2026
Further information: https://colalab.ai/
:quality(80))
:quality(80)){kind=icon}
:quality(80))