Research Fellow in Health Data Science (Machine Learning) — NeoShield Programme

Institution: London School of Hygiene and Tropical Medicine (LSHTM)
Location: London, UK (hybrid, minimum two days on-site per week)
Grade: 6 (£45,728–£51,872, inclusive of London weighting)
Duration: 24 months, full-time, with potential for extension
Funder: Wellcome Trust and Gates Foundation
Closing date: 21st April 2026
Start date: Immediately available

Want your models to reach the bedside?
We are recruiting a Research Fellow in Health Data Science to join NeoShield, a multi-country implementation research programme tackling neonatal sepsis and antimicrobial resistance across hospital neonatal units in Zambia and Malawi. This is a rare opportunity to take machine learning models from concept through to real-world clinical deployment in high-burden settings, working with large-scale routine clinical and microbiological datasets.

What you will build:
You will have end-to-end technical ownership of two production systems:

• A Clinical Decision Support Tool (CDST) for neonatal sepsis diagnosis, integrating clinical observations, point-of-care diagnostics and local microbiology trends to guide antibiotic treatment decisions.
• A real-time ward-level outbreak detection system using time-series analysis and anomaly detection to identify clusters of infection before they escalate.

The work spans the full pipeline: establishing and maintaining linked clinical and laboratory data sources, developing and validating supervised and unsupervised models, building dashboards and alerting tools, ensuring model interpretability and safety, and supporting integration into bedside digital workflows. You will travel regularly to Zambia and Malawi to work directly with clinical and laboratory teams on implementation, user testing and evaluation.

About NeoShield:
Neonatal sepsis remains one of the leading causes of newborn death globally, yet in many hospital settings in sub-Saharan Africa, infections are rarely diagnosed microbiologically, outbreaks go undetected until mortality rises, and antibiotic use is widespread but unguided by local data. NeoShield is an integrated programme spanning laboratory systems strengthening, antibiotic stewardship, genomic surveillance and machine-learning-driven clinical tools, led by LSHTM in partnership with the Malawi-Liverpool-Wellcome Trust (MLW) and the Zambia National Public Health Institute (ZNPHI). The programme is designed as a proof-of-concept with plans to scale across the region.

Who we are looking for:

• Essential: a postgraduate degree (ideally PhD) in machine learning, data science, statistics, computer science, epidemiology or another quantitative discipline; strong Python and/or R skills with version-controlled, reproducible workflows; substantial hands-on experience applying ML methods to real-world datasets; experience with temporal or time-series data and data pipelines; ability to work independently across a multidisciplinary, multi-country team; willingness to travel to Zambia and Malawi.
• Desirable: experience with healthcare or clinical datasets; model interpretability and bias assessment; deployment of models into production environments (dashboards, APIs, decision-support tools); familiarity with ethical and governance frameworks for ML in health.
  Infectious disease modelling experience is welcome but not required. We are looking for strong quantitative and engineering talent with a commitment to real-world impact.

What we offer:
The post-holder will play a leading role in scientific outputs and will be supported in pursuing future fellowships and funding opportunities. You will be affiliated with the LSHTM AMR Centre, MARCH Centre and DASH Centre, and will collaborate across all five NeoShield workstreams. Annual leave is 30 days plus discretionary wellbeing days, with pension scheme membership available. Visa sponsorship will be considered.

Apply here: https://jobs.lshtm.ac.uk/vacancy.aspx?ref=EPH-EPIH-2026-01-R

For questions about the role, contact James Henry Cross (james.cross@lshtm.ac.uk) or Eric Ohuma (eric.ohuma@lshtm.ac.uk).