Cloud Cost Optimizer API

An intelligent and scalable AI-powered Cloud Cost Optimizer API built with FastAPI, trained using both real and synthetic data, designed to recommend the most optimal AWS cloud resources based on user workloads, budget, and usage pattern. This project showcases practical implementation of Machine Learning, MLOps, Docker, and API key-based security, aimed to highlight skills relevant to modern backend and DevOps roles.

---

🚀 Features

• 🔍 Intelligent resource recommendation (EC2 instance type, storage, etc.)
• 📊 Trained ML model with EDA and rule-based fallback
• 🔐 API Key authentication using .env
• 🧪 Modularized codebase with separation of concerns
• 📦 Dockerized for seamless deployment
• ⚙️ Ready for deployment on Render, AWS EC2, or any container service
• 🧠 Designed to showcase MLOps and API development capabilities

---

📁 Project Structure

cloud-resource-recommender/
│
├── app/
│   ├── main.py              # FastAPI app entry point
│   ├── predictor.py         # ML-based prediction logic
│   ├── rule_based.py        # Fallback logic for unknown input
│   └── input_schema.py      # Request data validation with Pydantic
│
├── dataset/                 # Training datasets (real + synthetic)
│
├── models/                  # Saved ML model (.pkl or .joblib)
│
├── notebooks/               # EDA, preprocessing, training notebooks
│
├── requirements.txt         # Python dependencies
├── Dockerfile               # Docker image config
├── .dockerignore            # Ignored files during Docker build
├── .env                     # API Key and environment config
└── README.md

---

🧠 How It Works

1. Preprocessing & Training done in Jupyter notebooks:

   • EDA, feature engineering, and model training
   • Trained model saved to models/

2. API Layer (FastAPI):

   • Receives JSON input with workload requirements
   • Validates via Pydantic schema
   • Predicts optimal resource or falls back to rule-based logic

3. Security:

   • Requires API_KEY as header: X-API-Key

4. Deployment:

   • Fully Dockerized: docker build + docker run
   • Easily deployable to any cloud

---

🔑 API Usage (via Swagger)

After running the container:

http://localhost:8000/docs

Example Request:

Header:

X-API-Key: your_api_key_here

Body:

{
  "cpu": 4,
  "memory": 16,
  "storage": 100,
  "workload_type": "machine_learning",
  "budget": 150
}

Response:

{
  "recommended_instance": "t3.large",
  "estimated_cost": 145,
  "confidence": 0.87
}

---

🐳 Docker Instructions

Build the Docker image

docker build -t cloud-cost-optimizer-api .

Run the container

docker run -d -p 8000:8000 --env API_KEY=your_api_key cloud-cost-optimizer-api

---

✅ Requirements

fastapi
uvicorn
scikit-learn
pydantic
python-dotenv
joblib
numpy

Install with:

pip install -r requirements.txt

---

📈 Designed for Real-World Deployment and Use

• Shows end-to-end understanding of ML lifecycle
• Demonstrates practical FastAPI + Docker use
• Clean structure, good code practices, security-ready
• Deployable and scalable — fits real-world developer expectations

This project is tailored to showcase backend development, MLOps, and DevOps skills — emphasizing cloud efficiency, automation, and innovation in modern software solutions.

---

📌 Future Enhancements

• Add CI/CD with GitHub Actions
• Integrate with AWS Pricing APIs for real-time pricing
• Build optional front-end dashboard (React/Vite)
• Log requests and usage analytics (Prometheus/Grafana)

---

🤝 Contributions

This is an individual academic + placement showcase project. For feedback, open issues or connect with me!

---

🧠 Contact & Portfolio

• GitHub: [https://github.com/deepak5-ux]
• Docker Hub: [https://hub.docker.com/repositories/deepakb05]
• LinkedIn: [www.linkedin.com/in/deepak511]
• Mail: [deepakbalraj511@gmail.com]

---

"Optimizing the cloud, one API call at a time."