We underwrite small business loans using bank transaction data, accounting exports, and application signals. The current model is serviceable, but it does not explain decisions well enough for our credit team.
We need a senior data scientist to improve model performance while keeping interpretability front and center. You'll work with credit analysts, review bad approvals and false declines, and rebuild the feature set around cash-flow stability, seasonality, and repayment behavior.
Requirements
- Credit risk, underwriting, or fraud modelling experience
- Strong feature engineering and model validation skills
- Familiar with explainability and model monitoring
- Comfortable working with regulated decision-making workflows