Full-Stack Data Science Template

🚀 A comprehensive DS / MLOps project template with modern tooling, CI/CD pipelines, and starter files using Copier.

⚠️ Important: GitHub Template + Codespaces Only

This repository is designed exclusively as a GitHub template for use in GitHub Codespaces.

• 🎯 Use the "Use this template" button on GitHub to create your project
• 💻 Open in GitHub Codespaces for automatic environment setup
• 🚫 Do not use Copier directly - use the GitHub template workflow instead

Why This Approach?

This template uses a hybrid approach combining GitHub templates with Copier + initialization script because:

1. GitHub Codespaces Integration: Codespaces initializes with the main branch as root
2. Clean Project Structure: The init_project.py script transforms the template into a proper project structure
3. Automated Setup: Everything is configured automatically in the Codespaces environment
4. No Manual Configuration: Dependencies, tools, and environment are pre-configured

While it might seem redundant to use both Copier and an init script, this approach provides the cleanest development experience specifically for GitHub Codespaces users.

🚀 Quick Start

1. Create Your Project

1. Click "Use this template" on GitHub
2. Create your new repository
3. Open in GitHub Codespaces

2. Initialize Your Project

Once in Codespaces, run the initialization script:

This script will:

• ✨ Create proper project structure using Copier
• 🧹 Clean up template artifacts
• 🔧 Set up development tools (Ruff, Pyright, pytest, pre-commit)
• 📁 Move all files to the correct locations
• 🗑️ Remove initialization files

3. Verify Setup

📋 Prerequisites

• GitHub account with Codespaces access
• Python 3.8+ (automatically available in Codespaces)
• Copier installed

Note: All dependencies and tools are automatically installed during the initialization process.

Template Features

This template includes:

• Development Environment: Pre-configured with Python 3.12, Poetry/pip, and development tools
• Code Quality: Ruff for linting and formatting, pre-commit hooks
• Testing: Pytest setup with coverage reporting
• Feature Engineering: Optional Hamilton-based pipeline
• Model Training & Inference: MLOps-ready structure
• API Service: Optional FastAPI service for model serving
• Monitoring: Optional model drift detection and monitoring
• CI/CD: GitHub Actions workflows (optional)
• Deployment: GCP deployment configurations (optional)
• Documentation: Auto-generated project README

Updating an Existing Project

To update a project created from this template:

cd /path/to/your/project
copier update

This will apply any changes from the template while preserving your customizations.