Data Scientist

About the Role

Join our dynamic team at An Engineering company in Mumbai as a Data Scientist and play a pivotal role in transforming our customer-facing personal loan product. Collaborate with a skilled group of data professionals and internal stakeholders to leverage data-driven insights that shape our B2C offerings. This is an opportunity to significantly impact our credit assessment processes while advancing your career in a fast-paced environment.

Responsibilities

• Develop and maintain robust credit scorecards to enhance lending decisions.
• Engineer high-quality features from raw bureau data for model development.
• Design, implement, and monitor predictive models to assess portfolio risks effectively.
• Create signals to identify potential fraudulent activities within our systems.
• Collaborate with cross-functional teams to drive evidence-based product strategies.
• Analyze and interpret data to support risk management and limit policies.
• Deliver actionable insights to improve customer engagement and operational efficiency.

Requirements

• A minimum of 5 years of experience in data science or analytics is required.
• At least 3 years of specialized experience in digital lending or consumer credit is essential.
• Proven expertise in building and deploying credit scorecards is necessary.
• Experience in parsing and engineering features from raw bureau files is mandatory.
• Familiarity with fraud modeling, non-starter analytics, and limit management frameworks is expected.
• Strong proficiency in SQL and Python is a must for data manipulation and model development.
• Experience working with machine learning libraries such as pandas, sklearn, LightGBM, and XGBoost is required.

Nice to Have

• +Prior experience in shadow underwriting or champion–challenger frameworks is a plus.
• +Exposure to alternative data sources and SMS/device signals would be beneficial.
• +Familiarity with Random Forest and Statsmodels for advanced analytics is advantageous.
• +Knowledge of regulatory requirements in the lending space would be helpful.