Michael Luong

Built and launched a production-grade hybrid e-commerce platform (affiliate + dropship) with server-side analytics (GA4, Pixel, Clarity), managing 1,000+ SKUs via a unified storefront and React Admin Dashboard. Implemented automated product ingestion (500–1,000 products/day) using modular provider architecture to normalize heterogeneous data across eBay and AliExpress sources, achieving 99.9% uptime. Created a RAG-based AI shopping assistant with embeddings and MongoDB Atlas Vector Search for semantically-driven product discovery. Implemented a content & social distribution pipeline publishing to Facebook/Instagram via Meta Graph API, and built image-based search with Cloudinary for deduplication. Enabled real-time payments (Stripe/PayPal) with webhook-driven order confirmations. Developed a real-time customer-admin chat (Socket.IO) and a Tkinter-based dashboard for pipeline status.

Built and launched a production-grade hybrid eCommerce platform (affiliate + drop shipping) with server-side analytics (GA4, Pixel, Clarity) managing 1,000+ SKUs via a unified storefront and React Admin Dashboard. Engineered an automated product ingestion system processing 500–1,000 products/day, leveraging modular provider architecture to normalize heterogeneous data across the eBay and AliExpress data sources, achieving 99.9% uptime.

IMDb Movie Popularity Prediction: designed a random forest classifier to predict movie popularity, achieving 72% accuracy.