Com Tom

I’m a data-focused builder combining software engineering with machine learning to turn messy inputs into reliable, user-ready outputs. In my “Video Course RAG Assistant,” I designed a Whisper transcription-to-embedding pipeline and used FAISS for scalable similarity search, then implemented multi-topic query decomposition with Gemini-based answer generation grounded to exact video timestamps.

I’ve also built forecasting and recommendations systems—using SARIMAX for “Airline Route Demand Forecasting” and a hybrid association-rule + purchase-history approach for “Instacart Market Basket Analysis & Recommendation System.” Alongside this, I completed a “Web Development Intern — InternPe” where I built responsive React.js UI components and collaborated on debugging and version control with Git, and I’m actively strengthening my ML and data skills through coursework and hands-on projects.