Data Scientist

About the role

We are looking for a Data Scientist to join a clinical trial Feasibility team and develop the analytical capabilities that inform study planning, site selection, patient identification and study start-up. You will apply predictive modelling, causal inference and scenario modelling to help forecast trial performance and identify the practical levers available to improve it—from country and site selection through to enrolment and activation timelines.

The role combines rigorous hands-on data science with close business partnership. You will work with heterogeneous feasibility and operational data, reconcile entity relationships across multiple sources, and turn model outputs into clear, evidence-based recommendations for feasibility leads, study start-up teams and sponsors. You will be expected to work in a cloud-based MLOps environment and contribute models that are transparent, evaluated appropriately and fit for operational decision-making.

Key responsibilities

• Develop, evaluate and maintain regression, classification and ranking models supporting site and country prioritisation, enrolment forecasting and patient identification;

• Apply sound model calibration, validation and performance-monitoring approaches, especially where model outputs inform site rankings, patient identification estimates or other decisions with material delivery implications;

• Engineer features from site performance history, investigator and key opinion leader records, patient population and epidemiological data, competitive-trial landscape information, and operational systems such as CTMS and EDC;

• Build and maintain practical master-data linkages across sites, investigators, protocols and patients, resolving different identifiers and data structures before models are developed or used;

• Apply causal inference methods to distinguish the drivers of enrolment and activation performance from simple correlation;

• Develop scenario simulations that assess trade-offs across site mix, patient population assumptions and activation timing, creating sponsor-facing outputs that are both analytically robust and straightforward to interpret;

MUST-HAVE SKILLS

Applied Predictive Modelling in a Clinical Research or Life Sciences Context:

• Strong practical experience applying predictive modelling techniques in life sciences, clinical research or a closely related setting. You can build regression, classification and ranking models, selecting methods that match the decision problem and the underlying data.

• Exposure to clinical trial feasibility, site selection, study start-up, patient identification or enrolment data is strongly preferred.

Python, Machine Learning and Model Validation:

• Proficient in Python for data analysis and machine learning. You understand the full modelling workflow, including feature development, training, evaluation, calibration and validation. You can explain model performance and limitations clearly, particularly when models are used to rank sites or countries, estimate patient identification potential, or forecast enrolment;

Feature Engineering, Entity Resolution and Master Data Linkage:

• Hands-on experience engineering features from multiple heterogeneous data sources and resolving different identifiers for the same real-world entity. You can link sites, investigators, protocols and patients across operational, external and reference datasets, creating reliable analytical datasets from otherwise fragmented information;

Causal Inference and Scenario Modelling:

• Understand causal inference methods such as uplift modelling, counterfactual analysis or intervention-effect estimation, and can apply them to operational questions. You can investigate what changes enrolment or activation timelines, rather than reporting correlation alone. You are also able to build or contribute to scenario models that evaluate what-if choices across site mix, patient population assumptions and activation timing;

Cloud MLOps and Lakehouse Data Environments:

• You are comfortable working in a cloud-based MLOps environment and using modern data-platform patterns to access and transform analytical data.

• Experience with Azure Machine Learning, MLflow, lakehouse architectures or Delta Lake is particularly relevant. You can work effectively with shared data assets and contribute to reproducible, maintainable modelling workflows;

Communication and Business Translation:

• Translate complex analytical outputs into concise, evidence-based recommendations for non-technical stakeholders. You are comfortable explaining site rankings, enrolment forecasts, patient identification estimates and scenario comparisons to feasibility leads, study start-up teams and sponsors, making uncertainty and key assumptions clear;

GOOD TO HAVE — NOT A BLOCKER

• Direct experience with feasibility, country selection or site-selection data — useful for accelerating understanding of the business context;

• Experience with CTMS, EDC or comparable clinical operations data — valuable for working effectively with operational sources;

• Familiarity with investigator, key opinion leader, epidemiological or competitive-trial datasets — beneficial for developing high-quality feasibility features;

• Experience building sponsor-facing analytical products — helpful when communicating trade-offs and recommendations to external audiences;