Dec 2023 – Dec 2024 · 1 yr · Remote

• Part of the test design and measurement team that runs experiments and builds machine learning models, with focus on improving customer engagement and increasing revenue.

Jun 2023 – Aug 2023 · 2 mos · Atlanta, Georgia, United States · Hybrid

• Worked as data science intern in Test Design and Audience Measurement team. Completed following projects:
• Time-of-day optimization model: Developed a classification model that produces ranking of optimal promotional email delivery time of the day resulting in projected incremental revenue of apprx. $20M per year. the project involved refining the problem statement, selection of methodology, data engineering and model development.
• Propensity model: Built customer propensity model for oven range product that observed 60% recall rate achieved by targeting 20% customers

Aug 2023 – Dec 2023 · 4 mos · Atlanta, Georgia, United States · On-site

• Working as a Graduate Teaching Assistant for CSE6040: Computing for Data analysis

Jan 2023 – May 2023 · 4 mos · Atlanta, Georgia, United States · On-site

• Worked as Graduate Teaching Assitant for ISYE6501: Introduction to Analytics Modeling

Aug 2021 – May 2022 · 9 mos · Mumbai, Maharashtra, India

• Worked as an Associate in the Digital Fraud Risk Strategy team in consumer banking. Primary responsibilities fraud analysis, strategy creation, strategy deployment, and performance evaluation for RTP( Real-time payments) products.
• Reduced incoming fraud claims by almost 75% for RTP products.
• Deployed rule-based strategies to stop fraudulent transactions by detecting ATO(Account Takeover) patterns on the customer's account.

Jul 2020 – Aug 2021 · 1 yr 1 mo · Pune, Maharashtra, India

• Worked as a Model Developer in a 5 membered team tasked with building a suite of ML models for real-time authorization decisions for credit cards.
• Developed transaction risk scores that outperformed the industry benchmark risk scores.
• Designed and developed the framework for Data Creation and Feature Engineering, Model Selection and Optimization, and Model Evaluation.

Jul 2019 – Jul 2020 · 1 yr · Gurgaon, India

• Credit Card Fraud Strategy Refresh
• Developed an interactive Tableau dashboard for refreshing 50+ fraud strategy rules by changing thresholds of continuous variables
• Compared the performance of the rules using metrics such as Hit Rate, Catch Rate, FPR, Decline Rate, $ Captured Fraud
• Presented the impact of the new strategies on the portfolio level, successfully registering an incremental benefit of ~ $100K Month
• Operational Analytics for Fraud Prevention Team
• Generated a staffing schedule for optimal resource utilization and maximum alert coverage using a histogram of daily alert volume
• Presented the benefits of using Precision and FOR in contrast to Catch Rate for benchmarking operator’s performance
• Highlighted missed $ opportunity of no holds through metrics like % increase in holds and their effectiveness using Confusion Matrix
• Created dashboard used to monitor work status of alerts, the effectiveness of alert queues and operator utilization
• First Party Fraud Abuser Model
• Worked on POC to create a fraud abuser model to capture first-party fraud.
• Engineered aggregated features for first-party fraud abuser model such as inflow/outflow, credit utilization, big-ticket transaction, merchant history etc
• Established data creation pipeline on Pyspark and automated the refresh process using Crontab

Jul 2018 – Jul 2019 · 1 yr · Gurgaon, India

• Tableau Report for Card Fraud Reporting
• Created a tableau dashboard for monitoring fraud across Cards Portfolio
• Reported overall fraud trends, fraud declines,$ losses, industry benchmarks and emerging trends in fraud
• Analyzed data and provided insights across different categories such as LOB, fraud type, Wealth Groups and AMG clients
• Automated the data creation process to reduce report generation time by 50%
• Created Tableau dashboard for monitoring Zelle transactions using metrics like # transactions, $ flow, distinct accounts, $ buckets etc
• Data Engineering for Fraud Reporting
• Created and established a data pipeline for migration of fraud reporting tables from Hive to Pyspark (Data Frame API)
• Optimized and automated the process by scheduling Pyspark jobs using Crontab, reducing the execution time by ~90%