🛡️ Network Security ML Pipeline with FastAPI, MLflow, Docker, and AWS EC2

🧩 About this Project

This project combines machine learning and cybersecurity to detect phishing and malicious network traffic automatically.
What began as a small experiment grew into a complete end-to-end MLOps system — built with FastAPI, MLflow, MongoDB, Docker, GitHub Actions, and deployed on AWS EC2.

The pipeline can train itself on raw data, validate and transform it, select the best-performing model, and serve real-time predictions via a FastAPI endpoint.

---

🚀 Key Features

• End-to-end ML workflow: ingestion → validation → transformation → model training → deployment
• Trains multiple models automatically (Logistic Regression, KNN, Decision Tree, Random Forest, AdaBoost, Gradient Boosting)
• Evaluates models using F1-score, Precision, and Recall — tracked in MLflow via DagsHub
• FastAPI interface for live predictions and retraining
• MongoDB integration for secure data handling
• Containerized with Docker and automated CI/CD via GitHub Actions
• Deployed on AWS EC2 for real-time cloud inference
• Custom logging and exception handling for production reliability

---

🧠 Tech Stack

Languages & Frameworks: Python, FastAPI
ML Libraries: Scikit-learn, Pandas, NumPy
MLOps & Tracking: MLflow, DagsHub
Database: MongoDB
DevOps Tools: Docker, GitHub Actions, AWS EC2, dotenv