What are the responsibilities and job description for the Senior Data Scientist position at Cashera?
Please do not apply unless you have a background in financial services
Senior data scientist - Credit models
Cashera Instant is building the AI underwriting engine that will power our advance and credit line products. We are looking for a Senior Data Scientist with deep experience in credit model development to design, build, and continuously improve the decisioning system that determines who we lend to, how much, and at what price.
You will report to the Head of Credit and work closely with the CTO and Data Engineer. This is not a general ML role — we are looking for someone whose specific expertise is credit model science using financial transaction data.
Responsabilities:
- Build and maintain our core credit scoring and probability-of-default model for SMB advance and credit line products
- Ingest and engineer features from open banking data (bank transaction history, cash flow patterns, payroll cadence) and accounting APIs
- Design the pricing model that risk-adjusts advance cost based on predicted default probability
- Monitor model performance: accuracy, discrimination (GINI/AUC), calibration, and population stability over time
- Build the champion/challenger framework that allows continuous model improvement without disrupting live decisions
- Produce model documentation for regulatory review and for our institutional funding partners
- Partner with Compliance to test for fair lending (ECOA) compliance in model outputs
- Design early warning indicators that flag portfolio deterioration before it reaches charge-off
What we need:
- Has built a credit scorecard or probability-of-default model before not just classification models in general
- Expert in: logistic regression, gradient boosting (XGBoost, LightGBM), survival/hazard models for time-to-default
- Has worked with financial transaction data: bank statements, ACH flows, cash flow time series
- Understands ECOA and fair lending implications of ML credit models in the US
- Can make models interpretable, you will need to explain decisions to borrowers, regulators, and the credit committee
- Python stack: scikit-learn, XGBoost, pandas, numpy — the standard for credit modelling
- Background at a fintech lender, credit bureau (Experian, LexisNexis, FICO), or bank analytics team
Nice to have
- Experience building models on open banking / Plaid data specifically
- Has built seasonal adjustment models for businesses with non-linear revenue patterns
- Familiarity with the CFPB's small business lending data requirements (1071)
- Has built or validated models under SR 11-7 (model risk management guidance)
- Experience with MLflow or similar model tracking and deployment tooling