At Bluestock, I developed Python REST API components for an IPO financial data platform, implementing filtering, search, and sorting across 50K+ records. I also designed normalized PostgreSQL schemas for IPO, valuation, and analyst data.
On ModelMesh, I built provider-specific adapters that unify access to multiple AI model providers behind a versioned REST API. I added request validation, bounded timeouts, logging, and reliability metrics, and configured automated tests and GitHub Actions CI.
I built a RAG data pipeline combining ChromaDB vector search with PostgreSQL queries for a multi-agent financial intelligence platform. I also evaluated a Corrective RAG pipeline that reduced hallucinated LLM responses by 35%, and built a legal clause classifier that reached 64.49% macro-F1.
