Overview
Role: Applied AI Data Scientist
Location: Leeds, LS15 8GB. Hybrid schedule, 1-2 days a week in the office.
Salary: £60,000 - £75,000 per annum + up to a 10% annual discretionary bonus and extensive benefits.
Contract type: Permanent
Employment type: Full time
Working hours: Monday – Friday, 37.5 hours per week. Core hours 09:30 – 16:00; you can work around these to suit you.
We are the nation\’s largest online pharmacy, with 25 years\’ experience and over 1.8 million patients supported in England for NHS prescriptions. We are Great Place to Work certified and a certified B Corp, reflecting our focus on colleague experience and social/environmental responsibility. Our people are fundamental to our success as we pursue a patient-centric, digital healthcare vision. We are committed to maintaining a positive, open and honest working environment for all.
Role summary: Build and operate machine learning models underpinning Pharmacy2U\'s medication management products. Work with rich temporal and behavioural patient data to address problems with direct patient impact, including predicting medication need and identifying risk of non-adherence. This is a hands-on role within a small, high-impact team where models are productionised and embedded into patient-facing services.
Responsibilities
- Design, build, validate, and document machine learning models for medication behaviour, including adherence risk and medication synchronisation.
- Engineer temporal and behavioural features from prescription ordering patterns, cycle data, and adherence signals.
- Apply rigorous evaluation approaches, including cross-validation, calibration analysis, and fairness assessment across patient cohorts.
- Analyse large-scale medication ordering data to identify opportunities for new or improved AI-driven capabilities.
- Assess and communicate the clinical and commercial value of modelling approaches to support prioritisation and business cases.
- Collaborate with clinical stakeholders to define safety rules, constraints, and appropriate model usage in patient-facing contexts.
- Work with MLOps and engineering partners to package and deploy models into production environments (e.g., Azure ML).
- Define and support model monitoring, including performance baselines, drift detection, and retraining criteria.
Qualifications
- Demonstrated experience applying machine learning techniques, including classification, regression, and ensemble methods (e.g., XGBoost, LightGBM, random forests).
- Proficiency in Python for applied ML and analysis (pandas, scikit-learn, NumPy, matplotlib/seaborn).
- Experience engineering features from temporal, behavioural, or sequential data.
- Comfortable using SQL to explore and extract data from large relational databases.
- Experience working with large-scale tabular datasets (millions of records).
- Working knowledge of model interpretability and explainability techniques (e.g., SHAP, feature importance).
- Experience with robust model evaluation practices, including cross-validation, calibration, class imbalance, and metrics beyond accuracy (precision, recall, F1, AUC).
- Ability to communicate technical results clearly to non-technical stakeholders and document models for reuse and production.
- Background in applied data science or ML roles, with familiarity with regulated/healthcare contexts, cloud ML platforms, survival/time-to-event methods, and collaborative development practices (desirable).
What happens next?
Please click apply. If we think you are a good match, we will be in touch to arrange an interview.
Applicants must prove they have the right to live in the UK.
All successful applicants will be required to undergo a DBS check.
Unsolicited agency applications will be treated as a gift.
#J-18808-Ljbffr