How souhaib-ia/recommendation-engine works
Project Type and Purpose
• This is the backend microservice of a full-stack Movie Recommendation System. It exposes a REST API (Flask) that computes content-based movie suggestions (cosine similarity) and persists/reads data from a Supabase (PostgreSQL) database. It is deployed as an Azure Web App and packaged via GitHub Actions.
File Structure
• Top-Level Files:
• serve.py (entrypoint for Flask)
• requirements.txt (Python deps)
• Procfile (Azure/Gunicorn startup)
• .github/workflows/azure-webapps-python.yml (CI/CD to Azure)
• Implied Directories (from README):
• utils/ (data processing, feature extraction)
• models/ (cosine similarity, pre-trained artifacts)
• artifacts/ (precomputed matrices, uploads to Azure Storage)
README for Additional Insights
• Technology stack: Flask, Scikit-Learn, Supabase, Azure Web Apps
• API Endpoints: /recommend
• Deployment: GitHub Actions → Azure Web App, uses Gunicorn via Procfile