Projects

Selected work across NLP, ML systems, and analytics

A Scalable AI-Powered Receipt Intelligence System

Product Strategy Cross-Functional Leadership Roadmap & Execution KPI Design

Led the product strategy and cross-functional delivery of an in-house AI-powered receipt intelligence platform that replaced a high-cost third-party vendor. Delivered multimillion-dollar annual savings while supporting 10M+ users and 600+ brand partners.

Tech Lead & Product Owner → ML Technical Deep-Dive →

Designing a Revenue-Aligned Product Matching System at Scale

Revenue Strategy Diagnostic Analysis Hybrid ML Architecture Precision-Recall Calibration

Reframed product matching from an accuracy problem to a revenue-aligned systems problem. Conducted structured failure analysis, identified Pareto distribution in unmatched items, designed a two-track strategy (catalog enrichment + hybrid retrieval architecture), achieving 45% coverage improvement across millions of daily line items.

Product Strategy & Analysis → ML Technical Deep-Dive →

User Engagement & Retention Analytics

SQL Python pandas A/B Testing Tableau Snowflake

Built a unified engagement and retention analytics framework for millions of active users. Designed composite engagement scoring, cohort-based retention analysis, churn prediction models, and the experimentation layer for A/B testing product changes.

Read more →

Fake News Detection

Python NLP Scikit-learn Data Visualization

Text mining, EDA, and predictive modeling on the Buzzfeed news dataset to identify patterns distinguishing fake from real news. 31k+ views on Kaggle.

Read more →