langchain-with-c1-python

LangChain with Thesys C1 Python Example

This repo is a simple example of a natural language to sql query processing using Langchain and Generative UI API C1 by Thesys. The backend is a Python server powered by FastAPI and LangServe, and it can answer natural language questions about the Chinook sample database.

Project Structure

• frontend/: (Placeholder) A directory for a potential frontend application.
• backend/: A Python-based backend that exposes a LangChain agent as a REST API.

How to Run

Backend

1. Navigate to the backend directory, create and activate a virtual environment, and install dependencies:

   cd backend
   python -m venv .venv
   source .venv/bin/activate
   pip install -r requirements.txt

2. Set up the environment variables and run the server:

   You'll need a Thesys API key to run this example. Export it as an environment variable before running the server:

   export THESYS_API_KEY="your-api-key"
   python main.py

   The server will start on http://localhost:4001. You can access the API documentation at http://localhost:4001/docs.

Frontend

1. Navigate to the frontend directory, install dependencies and run the UI:

   cd frontend
   npm install
   npm run dev

   The UI will be available at http://localhost:5173.

What This Application Does

This application demonstrates how to build an intelligent SQL query assistant using LangChain and Thesys C1. Here's how it works:

1. Natural Language Processing: Users can ask questions in plain English about the music store data
2. SQL Generation: The LangChain agent powered by Thesys C1 converts natural language questions into appropriate SQL queries
3. Database Execution: The generated SQL queries are executed against the Chinook SQLite database
4. Formatted Results: Query results are returned in a human-readable format

The Chinook database contains information about a digital media store, including:

• Artists, Albums, and Tracks: Music catalog data
• Customers: Customer information and purchase history
• Employees: Staff information and sales data
• Playlists: User-created playlists
• Invoices: Sales transactions and billing

Sample Queries to Try

Once you have the application running, you can try asking questions like:

• "How many albums does each artist have?"
• "What are the top 5 best-selling tracks?"
• "Show me all customers from Canada"
• "Which employee has made the most sales?"
• "What genres are available in the store?"
• "Find all albums by Led Zeppelin"
• "What's the total revenue for each country?"
• "Show me the longest tracks in the database"
• "Which playlists contain jazz music?"
• "What are the most popular media types?"