All projects
Buffett Intelligence
A research tool for querying Warren Buffett's shareholder letters with natural-language search.
Web app2023FastAPI, Python, React, OpenSearch

The problem
Decades of Berkshire Hathaway shareholder letters contain enormous amounts of investment reasoning, but finding a specific idea means scrolling through PDFs. This was a course project focused on making that archive queryable.
Approach
The backend, built in FastAPI, indexed the letters and served natural-language queries. My focus was the frontend: a search interface that returns letter excerpts with enough surrounding context to be useful, not just a bare snippet.
Outcome
A working prototype that turns free-text questions into ranked excerpts from the letter archive, built and presented as part of coursework.
Next project