Research Fellow (Biological Sciences)

Job Description

About the role
The AI for Science Gym builds bottom-up AI capability across NUS science and engineering. Its Discovery Gym helps researchers who already know the basics of data science and ML use AI to push the frontiers of their own fields.
As a Discovery Gym Lead you are an AI-enabled data scientist embedded in a scientific domain. You run the Gym for one or more departments — guiding a weekly small-group cohort of postgraduates and postdocs, and partnering with research groups to turn their instrument data into working AI/ML capability, then into teaching material the whole University can learn from.

Job description

• Weekly small-group peer-learning in your department(s): participants work through hard problems on their own data and models, apprenticeship-style. Onboard each new department as it joins.
• One-month Discovery Sprints with domain partners: assess dataset suitability, agree goals and provenance, deliver a working dashboard within week 1, then define research goals and go/no-go gates. Close each sprint by handing over skills and scripts, and extracting a pedagogical dataset for the Gym gallery.
• Feed the shared substrate: deposit each partner's data as a tokenised pedagogical challenge (useful provenance, key annotations withheld), and contribute domain-workflow demos that bridge domain scientists and AI scientists.
• Publish and compete: run sprints that yield a pedagogical preprint; take part in scrimmages against competing AI meta-harnesses.
• Join the monthly DGL methods-exchange and the weekly sync with the Architect and other gym leads.

Working expectations:

• 1–2 active sprints at a time plus 1–2 in support; consultation ≤40% of your week; active projects wrapped within ~3 months; at least one paper or proceeding per year.

Qualifications

Ph.D. in a Science or Engineering discipline (e.g. Physics, Chemistry, Biology, Materials Science, Chemical/Biomedical Engineering, Pharmacy, food science).

Skills:

• Strong data science and ML experience: end-to-end pipelines — wrangling, dimensionality reduction, clustering, labelling, supervised and unsupervised learning — and an understanding of where they break.
• Fluent scientific Python with a modern ML stack (PyTorch or JAX, scikit-learn, pandas, numpy) and reproducible, version-controlled workflows.
• Experience with messy real-world instrument data, and the judgement to tell a promising dataset from a hopeless one.
• Ability to ship fast and visibly — a useful interactive dashboard on unfamiliar data within a week.
• Excellent communication and teaching instincts; genuine fluency with AI-assisted development and the ability to teach it critically.
• A publication track record and the appetite to keep publishing.

Experience:

• Postdoctoral or industry experience applying ML in research.
• HPC or multi-GPU environments; containerised workflows.
• Representation learning, foundation/self-supervised models, or physics-informed methods.
• Building or evaluating agentic/LLM workflows.
• Mentoring, teaching, or running workshops and competitions.
• Breadth across more than one scientific domain.