SmartSecure: An AI-Driven Network Threat Analysis & Detection Framework

The rapid growth of network-based services has led to a significant rise in sophisticated cyberattacks that closely resemble legitimate traffic patterns. Traditional signature-based Intrusion Detection Systems (IDS) are increasingly ineffective against such evolving threats, particularly zero-day and behavior-based attacks. This paper presents SmartSecure, a lightweight AI-based real-time Intrusion Detection System that combines offline machine learning training with live packet sniffing for dynamic attack detection. The proposed system captures network traffic, aggregates packets into flows, extracts behavioral features, and classifies traffic using a supervised RandomForest model trained on benchmark intrusion datasets. A real-time backend API streams detection results to a web-based dashboard for visualization and alerting. Experimental evaluation demonstrates that the system effectively detects multiple attack categories including DDoS, brute-force, port scanning, botnet communication, and web attacks while maintaining low computational overhead. The results indicate that SmartSecure bridges the gap between academic IDS research and practical, deployable cybersecurity solutions.