Featured Projects

Mining 300,000+ Amazon Electronics reviews with VADER and BERT to classify sentiment, surface product failure patterns, and visualize brand-level satisfaction trends across five product categories.

Querying Met, Louvre, and V&A APIs with Python to quantify shifts in Islamic art acquisitions and metadata evolution across 48 years of institutional records, revealing acquisition trends post-9/11.

Formulating baggage carousel assignment as a JuMP/Gurobi mixed-integer program to minimize passenger-weighted wait times and test fairness-aware scheduling under airport operational constraints.