GummiGhost

Overview

GummiGhost is an advanced AI-powered cyber defense system designed to operate on two integrated layers of security. It provides network-level protection with web-based intrusion detection using machine learning, and system-level defense with endpoint-level protection using local AI agents. It runs silently in the background to protect users, networks, and systems with minimal impact on performance.

Key Features

• Network-Level Protection - Detects threats like DDoS and port scans in real-time
• System-Level Defense - Monitors file behaviors, email threats, and malware
• Intrusion Detection System (IDS) based on NSL-KDD dataset
• Phishing Detection using NLP classification techniques
• Malware Detection analyzing files using behavior logs and API call sequences
• Real-time Security Dashboard for monitoring and insights

Technology Stack

• Frontend: React with TypeScript, TailwindCSS, Vite
• Backend: FastAPI, Pydantic, Scikit-learn, XGBoost
• ML/DL Models: NSL-KDD, BERT, TF-IDF + Logistic Regression
• System Agent: Python/Rust for performance and portability

Architecture

• Web-based Interface - Intrusion Detection, Phishing Detection, Malware Insights
• Local Agent (System Layer) - Real-time Malware Detection, Email Monitoring, File Scanning