Jan 2021 – Nov 2022 · 1 yr 10 mos · New York, United States

• I lead multiple data science work streams at Lyft:
• Reducing fraud losses by deploying ML models for identifying risky customers
• Improving membership retention rates through A/B testing and experimentation
• Improving payment health for Lyft Bikes and Scooters
• Building theft and loss fraud models for Bikes and scooters
• Improving customer experience of Lyft bikes and scooter riders by improving reliability of rides
• Most of my day to day is spent on running A/B tests, building ML models and doing statistical analysis to improve outcomes of aforementioned workstreams

Jan 2019 – Jan 2021 · 2 yrs

• Working on providing financial inclusion to immigrants, students and other underserved communities with the power of alternative data sources.

Feb 2017 – Feb 2019 · 2 yrs

• Developed 3 production level identity matching models for 114 million customers of Capital One for smooth sign in and sign up process by using state of the art technology stack (H2O, pyspark, aww).

Sep 2016 – Dec 2016 · 3 mos · New York City Metropolitan Area

• Use Machine learning techniques for analyzing the climatic drivers of global forest fires and come up with early warning mechanisms.

May 2016 – Sep 2016 · 4 mos · New York City Metropolitan Area

•  Developed optimization tool in R for budgeting marketing spend for various Pfizer brands
•  Identifying Health care providers that have a potential to grow their branded RXing by using Random Forest methods in Python
•  Developed a comprehensive index to assess potential county hospitals, for expanding Pfizer’s business in China using Python

Sep 2014 – Aug 2015 · 11 mos · Gurugram, Haryana, India

• Advised a major UK bank in their Asset Management for Secured Motor Loan Product (annual bookings GBP 800 M+).
•  Led a team of three analysts to develop pricing models for auto-finance loans leading to GBP 4M+ incremental sales
•  Forecasted defaults in certain recession scenarios for auto-finance portfolio with over GBP 800M+ annual bookings
• White Paper :
• Wrote a white paper analyzing the product development capabilities of leading Competitors of EXL and suggested areas in which EXL could successfully launch new analytical products
• Training:
• Conceptualized and conducted a training on basic statistics for EXL new joinees

Sep 2013 – Sep 2014 · 1 yr · Gurugram, Haryana, India

• Program Manager for Credit Card Collections in US
• Performed a Cost Benefit analysis for a temporary Collections Hardship program in the US Market and recommended increasing the enrollments by 2x..
• Awarded ‘Blue Rewards’ by American Express for my efforts in improving the identification process for deceased customers in the Collections program

Jun 2012 – Sep 2013 · 1 yr 3 mos · Greater Kolkata Area

• Collections and Capacity Planning Analyst :
• Developed a new linear optimization model for calculating number of agents in call center by mathematical formulating the Cost-Revenue and the constraints to this problems
• Improved the segmentation of delinquent credit card customers by adding more relevant variables in the decision tree increasing the information gain by 1 perecent

May 2011 – Jul 2011 · 2 mos · Greater Bengaluru Area

• Conceptualized and developed a tool in Microsoft Excel(VBA) for managing human resources. Currently being used by the company for employee scheduling and deciding if recruitment is necessary
• Developed an Oracle object Repository which contains information of all past projects undertaken by the company

May 2010 – Jul 2010 · 2 mos · Greater Paris Metropolitan Region

• Developed a simulator for a sensor Network in Java where individual sensors follow Markov Decision Processes.
• Used the simulator for testing approximate methods to predict steady state configurations for Sensor networks.
• Simulator is currently being used by the university to undertake research in game theory and physics