Technology consultant @ Tropical Digital

github

PRIYADHARSHAN VADIVEL • Peerlist

Project Title: AI-Driven Adaptive IDS with Deception Engineering for Critical Infrastructure Description: Engineered a next-generation Intrusion Detection System (IDS) that integrates Generative AI with cyber deception techniques to provide a dynamic and scalable defense mechanism for critical infrastructure. Leveraged a Generative AI model (LLaMA 3) to analyze network patterns and autonomously create adaptive Snort rules, enabling the system to evolve its defenses against novel threats. Developed Python scripts to dynamically block malicious IP addresses in real-time by integrating with firewall rules, immediately neutralizing identified threats. Implemented deception engineering principles by automatically redirecting attacker traffic to a secure decoy honeypot for advanced behavioral analysis and threat intelligence gathering. Established a structured logging system using MySQL to record all security alerts, providing a robust data source for forensic analysis and ensuring future compatibility with SIEM solutions.

Use of Generative Al for generation of Snort 

KeyGuard is a Python-based security tool designed to detect keyloggers hidden within .deb package files before installation. It uses custom YARA rules to scan extracted scripts and identify suspicious patterns, helping users secure their Linux environments.

KeyGuard – Keylogger Detection for .deb Files

HookScan is a Python-based phishing detection tool that analyzes URLs to identify potential phishing threats. This tool was developed to support ethical hacking practices by automating early-stage reconnaissance in red team assessments.