Amazon Copilot

A vector-based search system for Amazon product data using Qdrant for similarity search and sentence-transformers for embedding generation.

Academic Project: This project was developed for the course "Recuperación de Información y Recomendaciones en la Web" at Facultad de Ingeniería, Universidad de la República. Course information: https://eva.fing.edu.uy/course/view.php?id=821

Features

• Semantic search for Amazon products
• Vector embeddings with sentence-transformers
• Fast similarity search with Qdrant
• Type-safe Python codebase
• Command-line interface

Quick Setup

Prerequisites

• Python 3.12+
• Docker

Installation

1. Setup Python environment with pyenv:

   # Install Python 3.12
   pyenv install 3.12.2
   
   # Set local Python version
   pyenv local 3.12.2

2. Install dependencies with uv:

   # Install uv if you don't have it
   pip install uv
   
   # Create virtual environment
   uv venv
   
   # Activate virtual environment
   source .venv/bin/activate  # On Windows: .venv\Scripts\activate
   
   # Sync dependencies with specific extras
   uv sync --extra dev --extra backend
   
   # Or sync all dependency groups at once
   uv sync --all-extras

3. Start Qdrant database:

   # Use Docker Compose (recommended)
   docker-compose up -d

4. Configure environment:

   cp .env.example .env

Basic Usage

1. Create collection and load data

# Create a collection
amazon-copilot create-collection amazon_products

# Load product data from CSV
amazon-copilot load-products data/Amazon-Products.csv amazon_products

2. Search for products

# Search using the CLI
amazon-copilot search-products "wireless headphones" --collection-name amazon_products

3. API Usage

Run the FastAPI development server:

uvicorn amazon_copilot.api.main:app --reload

Access the API documentation at http://localhost:8000/docs

Documentation

For detailed guides, refer to:

• Detailed Setup Guide - Full installation instructions
• Data Loading Guide - How to load and manage product data
• Search Guide - Advanced search options and filtering
• Development Guide - For contributors and developers
• Qdrant Setup Guide - Vector database configuration

Project Structure

amazon-copilot/
├── src/amazon_copilot/
│   ├── api/               # FastAPI implementation
│   │   ├── main.py        # API entry point
│   │   └── routers/       # API route definitions
│   ├── services/          # Business logic services
│   ├── qdrant_client.py   # Qdrant vector DB client
│   ├── cli.py             # Command-line interface
│   ├── schemas.py         # Data schemas
│   └── utils.py           # Utility functions
├── data/                  # Data directory
├── docs/                  # Documentation
├── pyproject.toml         # Project configuration
├── docker-compose.yml     # Docker configuration
└── README.md              # This file

License

MIT License