A custom LangGraph workflow that converts natural language questions into SQL queries using open-source LLMs via Ollama. Built with custom tools to work with models that don't support function calling.

Tech Stack: LangChain, LangGraph, Ollama (CodeGemma 7B), SQLite Features: Sequential workflow (list tables → get schema → generate SQL → execute → natural language response) Usage: Run notebook_b.ipynb and ask questions like "How many customers are from the USA?" Output: Returns structured response with SQL query, raw results, and natural language explanation Database: Chinook music store database with 11 tables (Album, Artist, Customer, etc.)