Vandna Mirok

●Personal Loan Credit Risk Scorecard (ETB segment) : Developed a regression-based application scorecard for existing-to-bank personal loan customers, achieving a performance uplift of 10 Gini points over the incumbent model and ~25 Gini points over the CIBIL score, significantly improving underwriting accuracy.

●Auto Loan Risk Scorecard: Built risk scorecards for both new and used car loan segments. Enhanced approval rates by 20% while maintaining portfolio risk levels, leading to an incremental business growth of approximately INR 400 Cr.

●Bank Statement Score Analysis: Assessed the impact of incorporating a bank statement scorecard alongside existing personal and two wheeler application scorecards. Findings indicated a potential profit uplift of INR 50–60 Cr annually, driven by optimized approval and bad rate trade-offs.

●Personal Loan NTB Model Deployment: Contributed to the deployment of a challenger model for new-to-bank personal loan customers. Led coding and validation of deployment logic,and collaborated cross-functionally to ensure seamless production rollout.

●Technologies used: Python, pyspark, SQL, SAS, Jira

• Migrated back-end and front-end of PayPal's Auto Withdrawal Self Service Tool to PayPal opinionated stacks.
• Incorporated additional input components and features into the website to enhance search specificity.Refactored traditional React components into functional ones, elevating the website's user interface.
• Technologies used: Spring Boot (Java), PP-React (JavaScript).