Skip to main content

Projects

Vardia - Cyber Threat Intelligence Platform

A cyber threat intelligence platform that pulls CVEs, zero-days, ransomware leaks, darkweb postings and security news into a single feed, then deduplicates, prioritises and matches them against the technology stack an organisation actually runs.

The problem it solves

A security team starts the day checking five different places: NVD for CVEs, ransomware.live for leaks, BleepingComputer for news, plus forums and Telegram channels to see whose data is being sold. Vardia merges those into one feed and answers the question that actually matters: “Does this affect me?”

URL Guard - AI-Assisted Malicious URL Detection

A self-training, locally hosted phishing and malicious URL detector that combines machine learning, rule-based analysis, threat intelligence and AI into a single decision engine.

The problem it solves

Most URL checkers either look only at a blocklist — blind to fresh attacks — or return a single black-box score with no stated reasoning. URL Guard does both, and adds the judgement of a SOC analyst on top: it reads every piece of evidence, resolves the contradictions between them, and produces a reasoned verdict with a confidence score.

Fake Extension Builder

This project is a lightweight utility developed to obfuscate .exe file extensions by inserting the Braille Pattern Blank character (Unicode U+2800) just before the .exe suffix. The primary objective is to disguise the true nature of executable files in environments where basic file extension spoofing is relevant for educational, research, or cybersecurity awareness purposes.

Ryuk Ransomware

This project was developed to safely simulate and analyze how real-world ransomware operates within a controlled environment. Written in C#, this simulation uses AES encryption to lock files and places itself in system startup, accurately modeling the behavior of actual ransomware and providing a reliable test environment.

Note: This project was created solely for academic and educational purposes and must never be executed on real systems.