Production-grade Corrective RAG (CRAG) system built with LangGraph, FastAPI, Pinecone, and React, featuring retrieval grading, query rewriting, web fallback, short-term memory, and evaluation against naive RAG using RAGAS.
A simple example of an AI agent and a LangGraph with AI agent that can understand natural language queries about planet masses and perform calculations in scientific notation. Built as a demonstration of basic AI agent architecture with skills, prompts, and testing.
A LangGraph-orchestrated AI agent integrating Model Context Protocol (MCP), LangChain ToolNode execution, Streamlit, AsyncSQLite checkpointing, asynchronous event-loop scheduling, token-level streaming, persistent thread state, and configurable OpenAI/Gemini LLM backends.
To associate your repository with the sqlitesaver topic, visit your repo's landing page and select "manage topics."