I turn complex data
into tools people can use.

Built an LLM/RAG travel assistant with 90.45% answer accuracy, automated DAU and conversion reporting with SQL and Python, and developed a PyTorch LSTM for seven-day DAU forecasting.

Queried AWS-hosted PostgreSQL data in DBeaver to precompute taxonomy paths across 60,000+ records; analyzed mouse-tracking sessions to find points of user confusion and guide interface improvements.

Structured information on 500+ GIS programs into a searchable guide; used student survey findings and contributor feedback to prioritize discovery features and coordinate a global volunteer team.

Analyzed user and launch data for an AI hearing aid; ran online split tests of email subject lines and ad-platform messaging to inform Canadian market positioning.