AI & LLM Research Project
Personal Development / Self-Study
- Built an AI calendar assistant for natural-language scheduling and availability analysis across multiple calendars.
- Implemented agentic workflows with LangGraph / LangChain and OpenAI APIs; added RAG + memory for contextual decisions.
- Backend in Python (FastAPI) with PostgreSQL (+pgvector) and Redis; used LangSmith + a CLI harness for evaluation and rapid local testing.
- Developed as an R&D Proof-of-Concept (PoC) during a career sabbatical to explore scalable LLM architectures.
Tech Stack: Python, FastAPI, LangChain, LangGraph, LangSmith, OpenAI API, PostgreSQL (+pgvector), Redis, Docker, CI/CD, RAG.