Core Objective
Independently build a point-of-application risk scorecard from scratch within a tight window. The ideal candidate must take raw, uncleaned data and deliver a fully validated, production-ready logistic regression model.
Hands-on Scorecard Build History
Must have personally built and deployed at least 2 end-to-end credit scorecards (consumer or commercial). We are looking for someone who writes the code and build the bins rather than a manager or analyst who only reviews the output.
Hands-On Python Execution
Advanced, fluent Python coder.
Must be comfortable writing custom data transformation functions and debugging logic live without relying on template scripts or AI coders.
Pragmatic Data Engineering
Strong SQL skills to ingest, merge, and clean messy, high-dimensional datasets independently in cloud environments (GCP/BigQuery preferred).
The Data Scientist will
- Independently prepare complex datasets, run exploratory analysis, and build predictive models, risk scorecards, and decisioning tools
- Drive commercial product innovation - identify market gaps, prototype algorithms, take concepts to market-ready products
- Design and optimise data pipelines integrating large volumes of disparate commercial data
- Translate data assets into actionable business strategy and long-term analytics roadmap
- Apply advanced statistical/ML methods to uncover patterns in high-dimensional data
- Solve cross-domain problems (commercial risk, business failure, fraud detection) with engineering, product, and strategy teams
- Communicate complex findings clearly to technical and non-technical stakeholders
- Maintain data quality, governance, validation, and regulatory compliance standards
- Stay current with cloud capabilities (primarily GCP) and modern analytical tooling
- Mentor junior data scientists and lead code/quality reviews
The ideal Data Scientist will have the following experience
- STEM degree (Master's preferred)
- Proven experience working in a Data Scientist or quantitative modelling role
- Extensive experience with commercial data assets (eg business registry, trade credit, or bureau data)
- Strong commercial data interpretation, auditing, and validation skills
- Python and SQL (Unix/Shell Scripting a plus)
- Foundational credit risk modelling/scorecard life cycle knowledge (sampling, WoE, scaling)
- Git and CI/CD workflow experience
- Hands-on cloud development experience (GCP preferred)
- Exposure to ML methods (XGBoost, Random Forests, Neural Networks) and traditional stats (Logistic Regression)
- Awareness of data security, governance, and model risk management standards
Nice to Have
- UK commercial lending/regulatory knowledge (PRA/FCA, Consumer Duty, Basel 3.1)
- Experience building/validating commercial credit scorecards
- Exposure to Open Banking, transactional data, bureau feeds, ESG data
- Track record of independently pitching and delivering analytical products