Project: Image Quality Assessment o Led team of 4 to design image quality framework for Auto Navigator, car aggregator platform. Utilized object classification and generic image quality techniques, achieving ~95% accuracy, enhancing customer experience Project: Vehicle Valuation Capability o Led cross-functional team to develop a ML based vehicle valuation model to predict real-time car sale price outperforming market leading 3rd party valuation provider by ~15% o Re-imagined valuation use case to transform loss mitigation process saving ~8-10MM savings across multiple use cases

Project: Collections Phone Channel Strategy o Revamped customer engagement strategy for 1MM+ delinquent accounts and restructured calling operations in US and Philippineso Leveraged a ML based phone confidence model and an intelligent account segmentation algorithm to maximize collections increasing annual savings of ~$4M through additional payments Project: Vehicle Recovery Strategy o Led a team of 3 to co-design an automated address sharing platform leveraging a ML based address confidence model to improve collateral recovery performance driving annual revenue ~$10Mo Developed a dynamic segmentation to manage malicious borrowers driving annual revenue ~$1Mo Designed dashboards to further optimize manual effort in the process saving ~$1M annual operational expense

- Leading an offshore team of 4 to build a retail bank product recommendation model using the KNN technique to target existing credit card customers. Created customer segments using K-means clustering to implement KNN efficiently and measured product profitability based on historical revenue trends to find out the next best retail product for each customer. - Identified movement of credit card balances to installment loan balances and created prediction model using logistic regression, using on-us and off-us bureau variables to identify cross-sell opportunity and new target population.

Existing Campaigns Optimization and Sticker Testing for a Fortune 500 US credit card issuer- Led an offshore team of 3 to design the A/B testing framework for a comparison of multiple card activation stickers aimed at promoting digital activation. Identified best performing sticker with a potential $2.5MM+ annual benefit.- Led an offshore team of 2 analysts for optimization of existing marketing campaigns promoting digital space and identify potential targeting volume and impact of upcoming campaigns.- Devised an approach to measure the overall impact of marketing and servicing campaigns deployed throughout the year by creating a synthetic control population using stratified sampling.

Paperless: Digital Statements Strategic Growth Plan for a Fortune 500 US credit card issuer- Formulated a 4-year digital statements plan at the portfolio and overall level, producing a $35MM+ operational expense saves. This facilitated the goal-setting process for the client.- Developed a paperless propensity model to identify segments with higher chances of paperless adoption using the decision tree. Subsequently, devised customized targeting strategies for these segments translating into $3MM expense save annually.

• Capacity Planning and Optimization- Analyzing the forecast data and planning the manpower for the particular week in Advance.• Productivity improvement projects under Amazon Customer Excellence System (ACES). • Managing a team of 150 associates, 3 Process Assistants and 4 Problem Solvers within the Fulfillment Centre. • Ensuring the KPIs are met and making sure the OSHA compliance and Kaizen processes are followed.

- Investigated the Fluid flow characteristics in Thin Slab Continuous Caster and study the extent of effect by parameters affecting fluid flow on the resulting quality of steel with the help of extensive experiments.- correlated the experimental data with the plant data with the help of least squares regression to create a mathematical model to relate different parameters and optimise the resulting quality.