PydanticAI-Koyeb

A FastAPI web application that uses PydanticAI with Ollama to process CSV files containing product data.

The application can read CSV files, query a PostgreSQL database for regional data, and apply currency exchange rates to provide comprehensive product analysis.

Features

• CSV File Upload: Upload CSV files through a web interface
• AI-Powered Processing: Uses PydanticAI with Ollama for intelligent data processing
• Database Integration: Queries PostgreSQL database for regional product data
• Currency Conversion: Applies real-time exchange rates (USD to EUR)
• Real-time Logging: View processing logs in real-time
• Background Processing: Handles file processing asynchronously

Prerequisites

• Python 3.8+
• Ollama installed and running
• PostgreSQL database
• Free Currency API key

Installation

1. Clone the repository

   git clone <repository-url>
   cd example-pydanticai

2. Install dependencies

   pip install -r requirements.txt

3. Set up environment variables Create a .env file in the root directory with the following variables:

   OLLAMA_MODEL=mistral-small:24b
   OLLAMA_BASE_URL=http://localhost:11434
   DATABASE_URL=postgresql://username:password@localhost:5432/database_name
   FREE_CURRENCY_API_URL=https://api.freecurrencyapi.com
   FREE_CURRENCY_API_KEY=your_api_key_here

4. Set up the database Ensure your PostgreSQL database has a table named regional_data with the following schema:

   CREATE TABLE regional_data (
       region VARCHAR(255),
       product VARCHAR(255),
       sales INTEGER,
       amount DECIMAL(10,2)
   );

Usage

1. Start the application

   uvicorn app:app --host 0.0.0.0

2. Access the web interface Open your browser and navigate to http://localhost:8000

3. Upload and process CSV files

   • Upload a CSV file containing product data
   • Click "Process File" to start AI-powered analysis
   • View real-time processing logs
   • The system will:
     • Read the CSV file
     • Extract product names
     • Query regional data from the database
     • Apply USD to EUR exchange rates
     • Display results in a formatted table

Project Structure

PydanticAI-Koyeb/
├── app.py              # FastAPI application with web routes
├── agent.py            # PydanticAI agent implementation
├── requirements.txt    # Python dependencies
├── static/            # Static files (CSS, JS)
├── templates/         # HTML templates
│   ├── base.html      # Base template
│   └── upload.html    # Upload page template
└── uploads/           # Uploaded files directory (created automatically)

API Endpoints

• GET / - Home page with file upload form
• POST /upload - Upload CSV file
• POST /process - Process uploaded file with AI agent
• GET /log/{filename} - Get processing logs for a file

Technologies Used

• FastAPI: Modern web framework for building APIs
• PydanticAI: AI agent framework for intelligent data processing
• Ollama: Local LLM inference server
• PostgreSQL: Database for regional data storage
• Jinja2: Template engine for HTML rendering
• Pandas: Data manipulation and CSV processing
• HTMX: Dynamic UI updates without JavaScript

Environment Variables

Variable	Description	Required
OLLAMA_MODEL	Name of the Ollama model to use	Yes
OLLAMA_BASE_URL	Ollama server URL	Yes
DATABASE_URL	PostgreSQL connection string	Yes
FREE_CURRENCY_API_URL	Free Currency API base URL	Yes
FREE_CURRENCY_API_KEY	Free Currency API key	Yes

CSV Format

The application expects CSV files with product data. The CSV should contain columns that can be used to identify products. Example format:

product,sales,amount
Product A,100,1500.00
Product B,75,1200.00
Product C,200,3000.00

Deployment

This application is designed to be deployed on Koyeb. The FastAPI application runs on port 8000 and can be easily containerized or deployed as a serverless function.

License

This project is licensed under the MIT License.