MySonar API

MySonar is a Mobile Application designed to enhance the experience of Sónar Festival attendees through personalized schedule recommendations tailored to each user's musical tastes and attendance preferences.

This is a FastAPI backend that provides API endpoints for the MySonar mobile application, featuring activity recommendations powered by AI embeddings.

This repository only contains the code of the Backend API. If you are interested in the frontend, check out the Mobile App repository.

Features

• FastAPI REST API for festival activities and tags
• AI-powered activity recommendations using embeddings
• Supabase database integration
• Automatic API documentation with Swagger/OpenAPI

Setup

Prerequisites

Make sure you have the following installed:

• Python (version 3.8 or higher)
• pip (Python package installer)
• Supabase account and project

Get Started

1. Clone the repository

   git clone https://github.com/adriablancafort/mysonar-api.git
   cd mysonar-api

2. Install dependencies:

   pip install -r requirements.txt

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

   SUPABASE_URL=your_supabase_url_here
   SUPABASE_KEY=your_supabase_service_role_key_here
   
   OPENAI_API_KEY=your_openai_api_key_here
   

Development

1. Start the development server:

   python3 -m uvicorn main:app --reload

2. The API will be available at http://localhost:8000

3. The interactive API documentation (Swagger UI) will be available at http://localhost:8000/docs

4. Alternative API documentation (ReDoc) will be available at http://localhost:8000/redoc

AI Embeddings

This application uses AI embeddings to provide intelligent activity recommendations. Before using the recommendation features, you need to generate embeddings for your data:

Generate Embeddings

Run these scripts to generate embeddings for activities and tags:

# Generate embeddings for activities
python3 -m scripts.generate_activities_embedings

# Generate embeddings for tags
python3 -m scripts.generate_tags_embedings

These scripts will:

• Fetch all activities and tags from the database
• Generate AI embeddings for each item's text content
• Store the embeddings back in the database for use in recommendations