Senior Data Scientist

What You’ll Do

• Build and enhance credit risk models across multiple portfolios and regions, including PD, LGD, EAD, and lifetime ECL frameworks.
• Develop gradient boosting models (LightGBM, XGBoost) with strong calibration, backtesting, and out‑of‑time validation practices.
• Design vectorized engines for forward‑looking lifetime ECL calculations, integrating macroeconomic scenarios and discounting methodologies.
• Perform advanced feature engineering using payment history, delinquency trends, bureau data, and transactional behavior.
• Manage end‑to‑end model development using MLflow for experiment tracking, version control, and registry, ensuring full reproducibility and auditability.
• Build monitoring frameworks and dashboards to track model stability, performance, and data drift, with automated alerting across markets.
• Create macro‑overlay models that incorporate economic indicators (e.g., unemployment, GDP, interest rates) to support scenario‑based credit loss projections.
• Contribute to fair value estimation and coverage analysis supporting debt sales, pricing strategies, and capital planning.
• Execute monthly production scoring by loading trained models, processing exposure data at scale in the cloud, and validating ECL outputs.
• Maintain thorough model documentation and support audits, regulatory reviews, and independent validation processes.
• Partner with Data Engineering to define feature requirements, validate pipeline outputs, and ensure timely, accurate model inputs.
• Communicate findings and model insights to senior stakeholders, including Finance leadership, auditors, and regulatory bodies.

Who You Are

• 3+ years of experience in Data Science, Quantitative Analysis, or Credit Risk Modeling.
• Strong Python skills (pandas, NumPy, scikit‑learn) for modelling, analytics, and production‑ready code.
• Hands‑on experience with gradient boosting tools such as LightGBM, XGBoost, or CatBoost.
• Solid grasp of statistical concepts including probability, hypothesis testing, regression, time series, and survival/transition modelling.
• Proficient in SQL, capable of writing complex analytical queries for feature extraction and validation.
• Skilled in the full model lifecycle—training, tuning, validation, deployment, and ongoing monitoring.
• Familiar with experiment tracking platforms such as MLflow, Weights & Biases, or similar tools.
• Strong communicator able to translate complex model behavior and limitations to non‑technical audiences.

Nice to Have

• Experience with credit risk modelling (PD, LGD, EAD), transition matrices, vintage/roll‑rate analysis.
• Knowledge of IFRS 9 / CECL principles, staging logic, forward‑looking adjustments, and macroeconomic overlays.
• Familiarity with interpretability techniques (SHAP, feature importance, PDPs).
• Experience with Bayesian optimization for hyperparameter tuning.
• Exposure to Numba or vectorized computation for high‑performance modelling.
• Understanding of fair value or pricing methodologies for consumer credit portfolios.
• Experience deploying models on cloud infrastructure (AWS S3, Batch, Docker).
• Background in fintech, banking, or consumer lending.