Assur'Aimant: CI/CD & MLOps 🚀

---

Assur'Aimant is a professional Django application for insurance premium prediction. This project builds on the initial application built, integrating a complete CI/CD pipeline and MLOps best practices to ensure quality, reproducibility, and continuous deployment of ML/AI models.

---

🖼️ Application Preview

---

📋 Table of Contents

• Context & Objectives
• Features
• CI/CD & MLOps Strategy
• Installation & Launch
• Project Structure
• Security & Quality
• Contributing
• Authors
• License

---

🎯 Context & Objectives

As a Data AI Developer, this project aims to:

• Automate integration, validation, and deployment of AI models (CI/CD, MLOps)
• Ensure quality through automated tests covering the entire model lifecycle
• Deploy the model reliably and reproducibly

---

🚀 Features

• 🔒 Secure authentication & profile management
• 🤖 Insurance premium prediction
• 🧪 Automated tests on data, pipelines, training, evaluation
• 🛠️ CI pipeline (linting, tests, Docker build)
• 🚢 Continuous delivery (CD) with Docker & automated deployment (on Render)
• 📱 Responsive UI (Tailwind CSS)
• 🗃️ Prediction history & appointment management

---

🔄 CI/CD & MLOps Strategy

1. Automated Tests

• Data validation (columns, types, missing values/outliers)
• Data preparation tests (cleaning, scikit-learn pipelines)
• Training tests (convergence, hyperparameters, performance thresholds)
• Evaluation tests (metrics, bias checks)
• Tools: pytest, pytest-cov, pytest-mock, flake8, black, mypy

2. CI Pipeline

• Linting, formatting, type checking
• Running unit & integration tests
• Automated Docker build
• CI status badge in the README
• Workflows: .github/workflows/ci.yml

3. Continuous Delivery

• Model packaging via Docker (Dockerfile)
• Versioning by tag or commit SHA
• Publishing to registry (Docker Hub, GitHub Container Registry)
• Automatic deployment to staging, production on tag/approval (Render)
• Monitoring & rollback

---

⚙️ Installation & Launch

Prerequisites

• Python 3.9+
• pip
• Node.js (for Tailwind CSS)
• Docker

Quick Setup

# Clone the repo
git clone https://github.com/MichAdebayo/brief_django_application.git
cd brief_django_application

# Virtual environment
python -m venv venv
source venv/bin/activate  # Linux/MacOS
venv\Scripts\activate   # Windows

# Python dependencies
pip install -r requirements.txt

# Frontend dependencies (for Tailwind)
npm install -D tailwindcss postcss autoprefixer
npx tailwindcss init -p

# Migrations & admin
python src/brief_app/manage.py makemigrations
python src/brief_app/manage.py migrate
python src/brief_app/manage.py createsuperuser

# Start the server
python src/brief_app/manage.py runserver

Run tests & local CI

flake8 src/
black --check src/
mypy src/
pytest --maxfail=1 --disable-warnings -v

Build & run Docker

docker build -t assur-aimant:latest .
docker run -p 8000:8000 assur-aimant:latest

---

🗂️ Project Structure

brief_django_application/
├── src/
│   ├── brief_app/
│   │   ├── manage.py
│   │   ├── brief_app/
│   │   │   ├── settings.py
│   │   │   ├── urls.py
│   │   │   └── ...
│   │   ├── insurance_app/
│   │   │   ├── model/
│   │   │   ├── static/images/web-screenshot-1.png
│   │   │   ├── templates/
│   │   │   ├── forms.py, models.py, views.py, ...
│   │   ├── theme/
│   │   └── ...
├── tests/
├── Dockerfile
├── requirements.txt
├── .github/workflows/ci.yml
├── README.md
└── ...

---

🛡️ Security & Quality

• Password hashing (PBKDF2)
• CSRF protection
• Input validation
• Automated linting, formatting, type checking
• Automated tests for the entire model lifecycle

---

🤝 Contributing

1. Create a branch:

   git checkout -b feature/my-feature

2. Commit:

   git commit -m "Add my feature"

3. Push:

   git push origin feature/my-feature

4. Open a Pull Request

---

👨‍💻 Authors

• Michael Adebayo
• Eliandy Rymer

---

📄 License

Project under MIT License.