An intelligent wardrobe recommendation system that uses AI to generate personalized clothing suggestions based on user preferences, style, and context. The system leverages OpenAI's API and a multi-agent architecture to provide detailed wardrobe recommendations with real product suggestions.
- Personalized Recommendations: Generates wardrobe suggestions based on user's style preferences, occasion, and needs
- Real Product Search: Integrates with Walmart's product catalog to find actual clothing items
- Structured Output: Organizes recommendations by category (tops, bottoms, outerwear, etc.)
- Detailed Information: Provides product names, prices, descriptions, and URLs for each item
- Styling Tips: Includes personalized advice on how to combine and wear the suggested items
- REST API: Offers a Flask-based API endpoint for easy integration
- Clone the repository
- Install dependencies:
pip install -r requirements.txt- Create a
.envfile with your OpenAI API key:
OPENAI_API_KEY=your-api-key-here
Start the Flask server:
python main.pyMake a request:
curl -X POST http://127.0.0.1:5000/api/wardrobe/recommend \
-H "Content-Type: application/json" \
-d '{
"prompt": "I need a minimalist wardrobe with neutral colors for a professional setting"
}'{
"theme": "Professional Minimalist Wardrobe",
"styling_tips": "Mix and match these versatile pieces...",
"tops": [
{
"name": "Classic White Oxford Shirt",
"price": "$45.99",
"description": "Crisp cotton oxford shirt...",
"product_url": "https://...",
"image_url": "https://..."
}
],
// Other categories: bottoms, outerwear, headwear, footwear, accessories
}The system uses a multi-agent approach:
- Product Search Agent: Finds real clothing items matching the user's requirements
- Style Advisor Agent: Provides fashion advice and styling tips
- Main Wardrobe Agent: Coordinates the recommendations and ensures coherent output
- Python 3.8+
- OpenAI API
- Flask
- Pydantic for data validation
- Colorama for formatted console output
- Async support with nest_asyncio