Designed and implemented AI agents using LangChain for healthcare, integrating local large models via Ollama and remote APIs (OpenAI, Bailian) for versatile medical intelligence.
Developed multi-turn conversation memory management with MongoDB persistence.
Constructed a retrieval-augmented generation (RAG) pipeline by vectorizing medical documents, creating custom data loaders and splitters, and setting up vector-based search for instant clinical knowledge access.
Exposed common mathematical operations to the AI agent for enhanced functionality.
