README.md

metadata

title: L'Atelier du Fer - Catalog Search
emoji: 🔍
colorFrom: indigo
colorTo: blue
sdk: gradio
sdk_version: 5.29.0
app_file: app.py
pinned: false
license: apache-2.0
short_description: Search app for L'Atelier du Fer catalog

L'Atelier du Fer - Catalog Search

This application allows users to search for items in the decorative ironwork catalog of L'Atelier du Fer. It uses artificial intelligence to understand textual descriptions and analyze images to find the most relevant items.

Features

For Customers

• Search for items by text description
• Search for items by image
• Combined search (text + image)
• Display of results as thumbnails with image, title, description, price, and link to the product page

For Staff

• Quick identification of an item from a photo
• Display of detailed item information (price, reference, etc.)

Technologies Used

• Python 3.13.3
• Gradio for the user interface
• OpenAI GPT-4.1 for text and image analysis
• MongoDB Atlas Vector Search for vector search
• LangChain for component orchestration

Local Installation

1. Clone the repository:

git clone https://github.com/your-organization/adf-chatbot2.git
cd adf-chatbot2

2. Install dependencies with uv:

uv venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate
uv pip install -e .

3. Create a .env file at the root of the project with the following variables:

OPENAI_API_KEY=your_openai_api_key
MONGODB_URI=your_mongodb_atlas_uri

4. Launch the application:

python app.py

Deployment on Hugging Face Spaces

1. Fork this repository on GitHub
2. Create a new Space on Hugging Face
3. Select "GitHub" as the code source
4. Select your fork of the repository
5. Add the following secrets in the Space settings:
   • OPENAI_API_KEY: Your OpenAI API key
   • MONGODB_URI: Your MongoDB Atlas URI

Project Structure

• app.py: Application entry point
• src/: Application source code
  • database/: Database connection module
  • image/: Image processing module
  • search/: Vector search module
  • ui/: Gradio user interface module
  • utils/: Utilities and constants
• tests/: Unit tests

Running Tests

To run all unit tests:

uv run -m unittest discover tests

License

This project is licensed under Apache 2.0. See the LICENSE file for more details.