Led a six-person team developing an internal AI knowledge platform designed to make fragmented engineering information easier to discover. The prototype brought together knowledge from systems such as Jira, Confluence, and Zendesk through ingestion, semantic retrieval, and an AI-assisted interface.
OutcomeFunctional prototype demonstrated during an internal hackathon. Production adoption and retrieval performance are not yet documented publicly.
Technologies
- Python
- FastAPI
- PostgreSQL
- pgvector
- Azure OpenAI
- React
- Vite
- Node.js
- MCP



