This project is a Retrieval-Augmented Generation (RAG) chatbot built using LangGraph, LangChain, and Google Gemini. It allows users to interactively query and analyze student data stored in a local SQLite database.
- Conversational interface to query student information
- Uses a tool-augmented language model (LLM + SQL)
- Executes real-time queries on a SQLite database
- Built with LangGraph for dialogue flow management
- Includes sample student data for quick testing
- Python
- SQLite
- LangGraph
- LangChain
- Google Generative AI (Gemini)
- Google Colab
| File / Section | Description |
|---|---|
DatabaseManager |
Handles SQLite connection and queries |
ask_students_db |
LangChain tool for querying student database |
chatbot_node |
Generates assistant replies using Gemini |
human_node |
Handles user input |
build_graph() |
Creates LangGraph flow with nodes and edges |
setup_database() |
Initializes and populates sample data |
main() |
Runs the full application |
- Open the notebook in Google Colab.
- Set your
GOOGLE_API_KEYusinguserdata.get()(already in code). - Run all cells.
- The assistant will greet you with: Welcome to the Student Database Assistant...
- Ask questions like:
How many students are studying Computer Engineering?Who has the highest GPA?Show me all married students in PhD.
List all students over the age of 25.What is the average GPA for Physics majors?How many single students are there?
- Important: Use
check_same_thread=Falsewhen connecting to SQLite to avoid thread-related errors. - A Google Gemini API key is required to use the assistant.
MIT License (feel free to modify or extend)