Pixel Poisoning: Hacking Generative AI
A fierce visual defense against generative AI — fight back with poisoned pixels, stealthy watermarks, and adversarial trickery.
Read MoreDedicated third-year IT student with expertise in Cybersecurity, VAPT, and Cryptography, backed by global certifications. Skilled in problem-solving and innovation, seeking a role to enhance organizational security through advanced assessment and defense strategies.
I am a passionate Cybersecurity professional with a strong focus on Vulnerability Assessment and Penetration Testing. My expertise extends to cryptography, secure coding practices, and network security.
Currently pursuing my BS in Information technology Engineering, I am constantly expanding my knowledge and skills in the rapidly evolving field of information security.
All in One Crypto Solution with a self-developed ShadowHash Algorithm, offering encryption, breach detection, password generation, hash creation, comparison and malware scanning across 28 webpages.
Robust cryptographic key generator built in Java, utilizing PBKDF2, XOR operations, and multi-layered encryption (AES, 3DES, Blowfish) to generate secure, unpredictable keys for encryption.
Adversarial image obfuscation tool that generates ghost images using noise injection, edge distortion, metadata poisoning, and watermarking to prevent unauthorized use in AI training.
A steganography tool enabling users to hide and extract secret messages from images, featuring command-line interface, interactive mode, and lightweight encoding/decoding. Available on PyPI.
Python-based reverse shell for ethical hacking and penetration testing, offering persistence, keylogging, screenshot capture, file transfer, and stealth remote control over compromised systems.
A collection of Bash and PowerShell scripts for automating AWS tasks, including CLI setup, S3, EC2, and DynamoDB management, streamlining cloud operations for Linux and Windows users.
Python-based port scanner that detects open ports on single or multiple IPs, supports custom port ranges, uses socket timeouts for efficiency, and provides color-coded output.
This project implements a state-of-the-art digit recognition system trained on the MNIST dataset, achieving >98% accuracy. It features:
Percentage: 88.8%
Percentage: 93.8%
EC-Council
CompTIA
Google- Coursea
Self
Self
Cisco NetAcad
Amazon Web Services
Postman
IBA — Quest
"Robotics and Cybersecurity Fundamentals: Understanding Robotics, Penetration Testing Tools and Attack Vectors" accepted for publication in Book "Robot Automation: Principle, Design and Applications" (CRC Press, Taylor & Francis, Scopus Indexed) on 11/02/2025
"Hacking Generative Artificial Intelligence: Data Privacy via Image Poisoning" submitted for publication in book "Blockchain solutions for securing IOT networks: Practical applications and case studies"
Published multiple cybersecurity and scripting-related articles on Medium, featured in renowned publications like InfoSec Write-Ups and System Weakness
A fierce visual defense against generative AI — fight back with poisoned pixels, stealthy watermarks, and adversarial trickery.
Read More
Learn how to exploit SQL injection vulnerabilities using SQLMAP, an automated penetration testing tool for database security assessment.
Read More
A comprehensive guide to hacking and securing Wi-Fi networks using Aircrack-ng, covering practical attack techniques, encryption vulnerabilities, and prevention strategies.
Read More