Description :

Job Title : Data Scientist Credit Risk Analyst (Fintech)

Location : Delhi

Experience : 1 to 5 Years

Employment Type : Full-Time

Education : B.Tech from Tier 1 Colleges preferred

About the Role :

We are looking for a hands-on Data Scientist Credit Risk to build and optimize risk models for digital lending products. In this role, you will work on real-time underwriting models, alternate data-based risk assessment, and portfolio risk monitoring in a fast-paced fintech environment. You will collaborate closely with product, risk, and engineering teams to design robust credit models that enable scalable and responsible lending.

Key Responsibilities :

Digital Credit Risk Model Development :

- Build and deploy Application Scorecards, Behavioral Scorecards, and PD models for unsecured lending products.

- Define and construct target variables (DPD-based default definitions such as 30+, 60+, 90+ DPD).

- Design observation and performance windows aligned with business objectives.

- Develop logistic regression and ML-based models for underwriting and line assignment.

- Implement WOE transformation, IV analysis, feature engineering, and segmentation strategies.

Model Validation & Governance :

- Perform end-to-end model validation, including :

- Out-of-Time (OOT) validation

- Cross-validation

- Sensitivity and stability analysis

Evaluate models using :

- KS, AUC-ROC, Gini

- PSI (Population Stability Index)

- Lift & Gain charts

- Conduct back-testing and monitor model drift in live portfolios.

- Prepare validation documentation and support audit/compliance requirements.

Real-time Risk & Portfolio Monitoring

- Work with engineering teams to integrate models into underwriting pipelines.

- Monitor portfolio performance, delinquency trends, and risk segmentation.

- Recommend cut-off optimization and risk-based pricing strategies.

- Identify early warning signals and improve collections prioritization models.

Alternate Data & Advanced Analytics

- Leverage fintech-relevant data sources such as :

a. Bureau data

b. Transactional data

c. Bank statement analysis

d. Digital footprint/alternate data

- Experiment with advanced ML models while ensuring explainability

Technical Skills :

- Strong proficiency in Python (Pandas, NumPy, Scikit-learn, Statsmodels)

- Experience in :

a. Credit Risk Modeling in fintech/NBFC

b. PD modeling & scorecard development

c. Target variable construction

d. OOT validation & performance monitoring

- Strong understanding of :

a. Logistic Regression

b. WOE/IV

c. KS, Gini, AUC

d. PSI & Model Stability

- Good SQL skills for large-scale data handling.

Domain Knowledge :

Understanding of :

- Digital lending lifecycle

- Underwriting frameworks

- DPD-based default definitions

- Risk-based pricing

- Portfolio risk management