Pitch prize · ai@cam Sciencepreneurship (2026)
Won a pitch prize for a preventive health AI concept focused on making prevention more personalised and scalable.
InfoClaire Coffey, PhD
Researcher & consultant · Responsible AI for health
I use machine learning and data science to improve the fairness and real-world impact of predictive models, particularly in preventive health.
What I do
01
Responsible & explainable AI for health
Evaluation for trust in real settings: interpretability, robustness, and fairness across groups.
02
Clinical prediction & preventive health modelling
Risk prediction using clinical and population data, with validation designed for real-world use.
03
Reproducible ML pipelines
End-to-end, reusable modelling workflows that make analyses easier to audit, maintain, and extend.
About me
I am a researcher and consultant specialising in responsible AI for health. My work uses machine learning and clinical data science to improve the fairness and real-world impact of predictive models, particularly in preventive health. I hold a Visiting Researcher position at the University of Cambridge.
I completed a PhD in Health Data Science at the University of Cambridge, supported by a studentship from Health Data Research UK, The Alan Turing Institute, and the Wellcome Trust. My doctoral research examined fairness in clinical prediction algorithms and polygenic risk scores. Following this, I was a research scientist at Helmholtz Munich, where I developed an explainable machine learning framework for environmental health and a reproducible population-data modelling pipeline.
Earlier in my career, I worked in AI R&D at a startup designing multi-objective optimisation algorithms for autonomous vehicles and earned two patents. I also hold an MPhil in Advanced Computer Science (DeepMind Scholar) from the University of Cambridge and a BSc in Computer Science from the University of Birmingham, including visiting placements at the Universities of British Columbia and Waterloo.
Selected highlights
More
Recent milestones and updates.
- Won a pitch prize at ai@cam Sciencepreneurship for a preventive health pitch (making prevention more personalised and scalable).Feb 2026 ai@cam
- Delivered seminar: “Are polygenic risk scores fair for cardiovascular disease risk prediction?” (Data Science for Health Equity).
- Paper published: “Combined clinical, metabolomic, and polygenic scores for cardiovascular risk prediction” (European Heart Journal).Dec 2025 Paper
- PhD graduation ceremony (University of Cambridge).Oct 2025 Update
- Preprint released: “Current polygenic risk scores are unlikely to exacerbate unfairness in cardiovascular disease risk prediction”.Sep 2025 Preprint
- Submitted final hardbound PhD thesis copy (thesis available online).Aug 2025 Thesis
- Presented poster: “AI-driven environmental epidemiology: a generalisable framework to predict cardiovascular health using fair and explainable machine learning” (Helmholtz AI).Jun 2025 Poster
- Passed PhD viva with minor corrections.Mar 2025
- Joined Helmholtz Munich as a Research Scientist (AI & Environmental Health).Feb 2025 Helmholtz
- Submitted PhD thesis: “Evaluating and enhancing cardiovascular disease risk prediction using algorithmic fairness”.Dec 2024
- PPIE report released: “Equitable, trustworthy and safe research in healthcare technology, data and artificial intelligence: an international dialogue”.Dec 2024 Report
- Poster accepted for AMLD EPFL 2025: Algorithmic fairness in CVD risk prediction models.Dec 2024 AMLD
Selected papers, preprints, reports, and patents.
- Current polygenic risk scores are unlikely to exacerbate unfairness in cardiovascular disease risk prediction.medRxiv preprint · 2025 Preprint
- Combined clinical, metabolomic, and polygenic scores for cardiovascular risk prediction.European Heart Journal · 2025 Paper
- Axes of prognosis: Identifying subtypes of COVID-19 Outcomes.
- Achieving Net Zero within the NHS: System-wide transition to greener, sustainable care.
- Vehicle Route Guidance.US patent · 2019 Patent Google Patents
- Target Speed Optimisation.
Invited seminars and conference talks.
- Data Science for Health Equity Seminar: Are polygenic risk scores fair for cardiovascular disease risk prediction?2025 Watch
- Helmholtz Munich Seminar: Improving Cardiovascular Disease Prediction: From Fair Algorithms to Environmental Insights.2025 Helmholtz
- International Day of Women in Statistics and Data Science (IDWSDS) Conference: Data Science for Health Equity: Algorithmic Fairness in Cardiovascular Disease Risk Prediction.2024 IDWSDS
- ECML PKDD: Understanding and Improving Algorithmic Fairness in Cardiovascular Disease Risk Prediction.2024 ECML PKDD
- HDR UK Cambridge Community Meeting: Understanding and Improving Fairness in Medical Risk Prediction.2023 HDR UK
- HDR UK Doctoral Immersion Week on Fairness in healthcare data modelling: Algorithmic fairness in cardiovascular disease risk prediction.2023 Course
- AMIA Annual Symposium: Axes of prognosis: Identifying subtypes of COVID-19 Outcomes.
- HDR UK Bimonthly Webinar: Running and completing a PhD in the time of COVID (with Peter Diggle).2021 Webinar Peter Diggle
- UCL MedTech Conference: Sharing my experiences in MedTech.2021 UCL MedTech
I’m keen to do more science communication work — public engagement is a core part of doing useful research.
- Cambridge Festival: Artificial Intelligence — With great power comes great responsibility.
- Cambridge Festival: Fairness in medical risk prediction algorithms.
- Blog posts.
- Artificial Intelligence: Theory, Responsibility, and Sustainability (co-designed and delivered).2023
- MPhil Population Health Sciences — Teaching assistant (Advanced Statistics for Epidemiology).2023
- High school Computer Science — GCSE/A-Level tutoring and in-school teaching (years 7–13).2017–2022
- Statistical Methods for Health Equity organiser (seminars, reading group, research; yearly symposium).
- Journal manuscript reviewer (collaborative lab reviews).Science · European Heart Journal
- Reinforcement learning for healthcare (offline RL; multimorbidity trajectory modelling).Supervised by Chris Yau Chris Yau
- Fairness and the bias–variance trade-off (MPhil thesis; criminal recidivism data).Supervised by Neil Lawrence Neil Lawrence
- Multi-objective journey optimisation (BSc thesis; autonomous vehicles).Supervised by Dave Parker Dave Parker
- Risk communication: evaluating tools for presenting disease risk to patients.Project led by Owen Taylor Project Owen Taylor