Dec 2019 – Mar 2022 · 2 yrs 3 mos · Sunnyvale, CA · On-site

• Experimentation and Feature Launches
• Successfully launched 10+ features, generating incremental $30MM in annualized revenue. Key initiatives included introducing standardized login and membership modules on the Sam’s Club website through feature enhancements, platform migrations, and multi-variant experiments.
• Anomaly Detection
• Developed real-time anomaly detection algorithms using Median Absolute Deviation (MAD) in Python, complemented by visualization dashboards for executive and product management insights.
• Campaign Performance Optimization
• Implemented a synthetic control method to measure campaign performance, enabling precise lift analysis and actionable insights for improved marketing effectiveness.
• Comparative Analysis
• Implemented both Bayesian and frequentist methodologies to optimize experiment designs and conducted difference-in-differences analysis for post-launch impact evaluation.
• Leadership and Team Management
• Led login and membership digital optimization channels, overseeing a team of 2 data scientists with a focus on mentorship and professional growth.

Jan 2018 – Jan 2019 · 1 yr · San Francisco Bay Area · On-site

• Predictive User Feedback System
• Collaborated with product managers and customer support teams to develop a predictive user feedback system, leveraging topic modeling and text classification techniques such as multinomial Naive Bayes, Word2Vec, and word embeddings. This system provided actionable insights that improved consumer retention and engagement.
• Propensity Scoring Model
• Designed an imbalanced scoring model to estimate a consumer's propensity to purchase a house using activity and demographic attributes, enhancing customer satisfaction by effectively quantifying ROI.
• Customer Churn Prediction
• Developed a customer churn prediction model utilizing gradient boosting and bagging techniques, resulting in a 13% year-over-year increase in retention rates.
• Dynamic Pricing Model
• Built a multi-variant testing model using Thompson sampling to dynamically adjust pricing across markets, optimizing revenue generation and market competitiveness.

Jan 2017 – Jan 2018 · 1 yr · San Francisco Bay Area · On-site

• Customer Segmentation and Campaign Optimization
• Developed classification algorithms to segment customers based on behavioral and predictive attributes, enabling targeted and effective marketing campaigns.
• Data Storytelling and KPI Metrics
• Created comprehensive data stories and business KPI metrics to track and showcase the growth of Visa Advertising Solutions.
• OKR Tracking and Visualization
• Designed advanced visualization tools to monitor and evaluate product OKRs, ensuring alignment with strategic objectives.
• Client Relationship Management
• Managed client expectations through proactive communication and delivery of tailored insights and solutions.

Jan 2016 – Jan 2017 · 1 yr · San Francisco Bay Area · On-site

• Real-Time Decisioning Chatbot Development
• Developed a real-time decisioning chatbot leveraging advanced natural language processing (NLP) techniques such as Word2Vec, topic modeling, and Multinomial Naïve Bayes to process unstructured data. The chatbot efficiently classified and resolved customer issues, streamlining query management and enhancing user experience.
• Cost Reduction and Scalability Enhancement
• By automating repetitive tasks and significantly improving resolution times, the chatbot reduced operational costs by 10% year-over-year. This solution also enhanced scalability, enabling the support team to manage increased query volumes effectively.
• Collaborative Filtering Recommendation System
• Developed an item-item and hybrid collaborative filtering recommendation system for the sales team, analyzing sales data and customer history to generate personalized recommendations. This approach enhanced cross-selling, increased sales productivity, and drove revenue growth.
• Global Team and Client Management
• Managed a distributed team across five geographies, gathering requirements, creating BRDs, and ensuring seamless stakeholder communication. Delivered client presentations on progress and insights, and prepared leadership-ready materials for executive reviews to align with strategic goals.

Aug 2013 – Jan 2016 · 2 yrs 5 mos · Greater New York City Area · Remote

• Regulatory Submissions Management
• Managed a team responsible for regulatory submissions of Citibank’s North America credit card portfolios, ensuring compliance with Basel III, FDIC, and CCAR requirements.
• Data Extraction and Reporting Framework
• Developed an automated data extraction and reporting framework in SAS, streamlining processes such as data integrity checks, report creation, and dashboard generation. This framework improved efficiency by 200% and enabled proactive identification of high-risk escalations.
• Risk Parameter Modeling
• Built Decision Tree (CART) models to calculate key risk parameters, including PD (Probability of Default), CCF (Credit Conversion Factor), EAD (Exposure at Default), LGD (Loss Given Default), and RWA (Risk-Weighted Assets). Collaborated closely with regulators to gather specifications and implement them effectively in models.
• Portfolio Credit Performance Analysis
• Analyzed credit performance across portfolios, preparing detailed reports and documents to communicate insights to Citibank management and auditors.
• Credit Card Authorization Strategies
• Worked with Citibank’s Policy team to develop credit card authorization strategies, addressing functionalities like over-the-limit transactions, delinquency management, and bust-out prevention.
• Fraud Detection Optimization
• Assisted Citi’s Fraud Detection team in synchronizing 800 authorization rules and 500 fraud rules, resulting in a more robust and secure platform.
• Default Prediction Models
• Collaborated with the Model Development team to build predictive models for customer-level credit defaults, enhancing risk assessment and mitigation strategies.

Jul 2012 – Aug 2013 · 1 yr 1 mo · Gurgaon,Haryana · On-site

• Sales Force Payout Methodology
• Devised methodologies to determine quarterly and annual payouts for the U.S. pharmaceutical sales force at GlaxoSmithKline. Successfully led three critical modules within 10 months of joining, including eligibility assessment, field competency evaluation, and goal and quota setting.
• Performance-Based Payout Model
• Developed a performance and sales-based payout model using clustering (K-means), multivariate, and bivariate regression techniques in SAS/STAT and SQL, enabling data-driven and equitable compensation strategies.
• Sales Target Prediction
• Predicted individual sales targets for medical representatives using time series analysis, enhancing accuracy in goal-setting and sales performance evaluation.