Product discovery built around real-world dietary and budget constraints.

Designed and implemented the ranking algorithm at the core of a FastAPI and scikit-learn backend, scoring and sorting products against budget and dietary constraints across four stores and twenty products. Wrote and passed sixteen automated tests to validate the ranking, search, and budget algorithm logic; extended a team usability study with eight participants into an independently engineered, deployed full-stack prototype. Integrated the Claude API as an optional in-app assistant, connecting natural-language queries to the underlying product search and ranking pipeline.

Scope of Work