Job Description

Role purpose

As a Credit Data Scientist, you’ll use data, feature engineering and experimentation to improve credit decisioning and portfolio performance across our lending products and markets. You’ll work end-to-end from data exploration through to production-aligned features, monitoring and impact measurement.

Key responsibilities

· Analyse customer, bureau, transactional and repayment data to identify drivers of risk, loss, approval rates and customer outcomes.

· Build and iterate credit risk features and model inputs (behavioural signals, affordability proxies, stability-tested transformations), partnering closely with senior modellers and engineering.

· Contribute to development and improvement of predictive models using modern machine learning approaches, with a focus on robustness, stability and deployability.

· Design, run and evaluate credit policy experiments (cut-offs, limits, pricing/risk trade-offs, segment strategies), including post-implementation reviews.

· Develop monitoring for model/policy performance and feature health (drift, stability, segment performance, data quality checks).

· Support portfolio analytics: vintage analysis, roll-rates, migration, early warning indicators, collections funnel analytics, and loss driver deep-dives.

· Work with Data/Engineering to improve data definitions, quality, lineage and reproducible pipelines; document feature logic and assumptions.

· Contribute to governance documentation (model inputs, feature catalogues, monitoring evidence, change logs).

Requirements

Required experience and qualifications

· 2–4 years in credit analytics / credit risk / lending data science (bank, fintech, lender, bureau, consulting).

· Strong Python and/or SQL skills and experience working with large datasets.

· Proficiency in Python or R for analysis and modelling.

· Solid grounding in statistics and predictive model evaluation (ranking performance, calibration, stability) and business impact measurement.

· Exposure to advanced machine learning concepts (e.g., ensemble methods, cross-validation, hyperparameter tuning) and an understanding of how to apply them responsibly in production settings.

· Clear communication skills with technical and non-technical stakeholders.

Nice to have

· Experience with bureau data, open banking/transactional data, device/behavioural signals, or alternative data.

· Familiarity with model monitoring, governance, and documentation practices in regulated environments.

· Exposure to cloud analytics stacks (e.g., BigQuery/Snowflake/Databricks) and version control (Git based).

Personal attributes

· Curious and pragmatic; focused on measurable outcomes.

· Comfortable working in detail and iterating quickly while maintaining quality.

· Collaborative and able to work across markets and time zones.

Reporting line and location

· Reports into credit analytics center of excelence.

· Location: Mumbai, India. With collaboration with in-country lending and credit risk teams.