A group of final-year Information Security Engineering students from the National University of Sciences and Technology (NUST) has created an advanced AI-powered cyber threat intelligence platform named Cyber Vigilant, earning them a top-three rank in their department and a coveted spot in the competition for the prestigious Rector’s Gold Medal.
Developed in collaboration with Ebryx (Pvt.) Ltd., Cyber Vigilant is a unified cyber defense system that combines artificial intelligence, deep packet inspection, and encrypted traffic monitoring to offer real-time threat detection and analysis. The student team, led by project leader Haiqa Hashmi along with Rahat Jan, Zoha Naeem and Muhammad Ali Naveed, worked under the academic supervision of Ms. Aimen Akif and industrial guidance from Ebryx’s Muhammad Talha Masood. Their combined effort and technical excellence have created a platform that resonates with the real-world needs of cybersecurity professionals.
The project stands out not just for its academic merit but for its robust, enterprise-level features, making it relevant for organizations navigating an increasingly complex threat landscape. At its core, Cyber Vigilant includes four integrated modules — an analysis dashboard, phishing email detection, AI-driven anomaly detection, and encrypted traffic monitoring.
The Analysis Dashboard offers real-time visibility into known web attacks. One of its standout features is color-coded alerting linked directly to the corresponding lines in system logs, allowing cybersecurity teams to pinpoint threats with precision. Moreover, the dashboard includes a global IP geolocation map that visualizes all current network connections, helping analysts understand potential threat origins across geographical locations.
The Phishing Email Detection module automates the process of identifying and analyzing suspicious emails. It uses URL and content inspection to flag phishing attempts and offers advanced features such as manual email header input and manual URL analysis. This module can detect spoofed sender addresses, embedded malicious links, and unusual language patterns—tools increasingly necessary as email-based threats grow more sophisticated.
The AI-Driven Anomaly Detection component tackles the complex issue of zero-day attacks—unknown threats that have no prior detection signatures. This module employs a hybrid machine learning approach: unsupervised One-Class SVM models to detect anomalies in new traffic and supervised Random Forest algorithms to classify known attack types. The system captures live traffic, converts it into structured data formats (CSV), and processes it through AI models, generating detailed reports that can be used for further threat analysis.
The final module, Real-Time Encrypted Traffic Monitoring, is a testament to the system’s depth. It inspects encrypted web traffic—an area often overlooked in conventional threat detection systems. Using deep packet inspection (DPI) even on secure connections, the platform can detect threats hidden within SSL/TLS layers and block them in real time via a built-in proxy.
With the increasing reliance on encrypted channels and the growing number of zero-day exploits, a system like Cyber Vigilant represents a meaningful step forward in academic contributions to cybersecurity. Its design aligns with the needs of both enterprise security teams and national security frameworks.
The student team expressed gratitude for the mentorship and collaboration that brought the project to life. “We’re incredibly proud of what we’ve achieved,” said team leader Rahat Jan. “Being shortlisted for the Rector’s Gold Medal is an honor, but more importantly, we hope our work contributes to the broader cybersecurity ecosystem in Pakistan.”
As cyber threats continue to evolve in complexity and volume, innovative projects like Cyber Vigilant reflect the growing talent and ambition of Pakistan’s next generation of security engineers.
Source: LinkedIn
