Senior Data Modeler

CreditAccess Grameen Limited; Bengaluru; On-site; INR 1,200,000 - 2,400,000; Full time

Job summary

CreditAccess Grameen Limited in Bengaluru seeks a Senior Data Modeller to design and build predictive models for lending decisions, including propensity and application scoring, with end-to-end ownership from feature engineering to deployment.

You will work with data analysts and risk, credit, or marketing stakeholders to translate business needs into model design, validate performance, and support monitoring and recalibration.

Qualifications

• 5-7 years of experience building statistical / machine learning models in a Bank or NBFC, with lending experience mandatory.
• Hands-on experience building propensity, cross-sell, risk, or application scoring models using techniques such as logistic regression, XGBoost, or other tree-based methods.
• Strong Python (or R) skills for model development, plus strong hands-on SQL for feature and dataset preparation.
• Experience modelling in at least one of: risk/credit, or marketing/cross-sell.
• Ability to explain model logic, drivers, and trade-offs clearly to business and risk stakeholders.

Responsibilities

• Design and build predictive models such as propensity / cross-sell models, application scorecards, or risk/delinquency models depending on domain fit.
• Engineer features from transactional, bureau, collections, or campaign data to feed model development.
• Apply and compare modelling techniques (e.g., logistic regression, XGBoost / gradient boosting, decision trees) to select the best-fit approach.
• Validate model performance metrics, stability checks, and back-testing and document model logic and assumptions for governance and audit.
• Partner with data analysts and Risk, Credit, or Marketing stakeholders (as aligned) to translate business objectives into model design choices.
• Support model deployment, scoring pipeline handoff, and periodic model monitoring/recalibration.
• Write and optimise SQL for feature extraction and model dataset preparation.

Skills

Python

SQL

R

Propensity modelling

Cross-sell modelling

Application scoring

Risk modelling

XGBoost

Decision trees

Logistic regression

Education

Bachelor's or Master's in Statistics, Economics, Engineering, CS

Tools

SQL

Job description

We are augmenting our Data Intelligence team with a Senior Data Modeller who can build statistical and machine learning models that directly support lending decisions: propensity, risk, cross-sell, and application scoring, depending on the candidate's strongest background. This role owns the model build end-to-end: from feature creation through to a validated, deployable model, working closely with data analysts and Risk, Credit, or Marketing stakeholders on model use and adoption.

Responsibilities:

• Design and build predictive models such as propensity / cross-sell models, application scorecards, or risk/delinquency models depending on domain fit.
• Engineer features from transactional, bureau, collections, or campaign data to feed model development.
• Apply and compare modelling techniques (e. g., logistic regression, XGBoost / gradient boosting, decision trees, and similar classification techniques) to select the best-fit approach.
• Validate model performance metrics, stability checks, and back-testing and document model logic and assumptions for governance and audit.
• Partner with data analysts and Risk, Credit, or Marketing stakeholders (as aligned) to translate business objectives into model design choices.
• Support model deployment, scoring pipeline handoff, and periodic model monitoring/recalibration.
• Write and optimise SQL for feature extraction and model dataset preparation.

Requirements:

• 5-7 years of experience building statistical / machine learning models in a Bank or NBFC, with lending experience mandatory (microfinance exposure not required).
• Hands-on experience building propensity, cross-sell, risk, or application scoring models using techniques such as logistic regression, XGBoost, or other tree-based / classification methods.
• Strong Python (or R) skills for model development, plus strong hands-on SQL for feature and dataset preparation.
• Experience modelling in at least one of: risk/credit, or marketing/cross-sell.
• Ability to explain model logic, drivers, and trade-offs clearly to business and risk stakeholders.

Good to Have:

• Experience with model validation, monitoring, or governance frameworks.
• Exposure to credit bureau data, collections data, or campaign/CRM data.
• Familiarity with loan lifecycle metrics (delinquency, vintage, roll-rate) or marketing metrics (funnel, attribution, CLV).
• Experience with dimensional/semantic data modelling (star schema or equivalent) as a supporting skill.
• Bachelor's or Master's degree in a quantitative discipline Statistics, Economics, Engineering, Computer Science, or related field.