Projects
Case studies of security tools and forensic automation systems.
A secure, zero-knowledge, real-time messaging platform using Python WebSockets and client-side OpenPGP encryption.
Problem
Standard messaging platforms often compromise user privacy through persistent message logging, metadata retention, and lack of verifiable end-to-end encryption.
Investigation
Researched cryptographic standards for browser-based encryption. Selected OpenPGP.js for robust client-side encryption. Analyzed WebSocket performance for real-time bidirectional communication.
Architecture
Client-side OpenPGP encryption with in-browser key generation. Python WebSocket server acting as a blind relay (zero-knowledge). Ephemeral session-based identity system.
Stack
Outcome
Delivered a fully functional secure messaging proof-of-concept with true end-to-end encryption, anonymous usage, and a cyberpunk terminal aesthetic.
Network Vulnerability Scanner
Developed a custom vulnerability assessment tool using Nmap and Metasploit. Identified 150+ vulnerabilities across 50+ endpoints, reducing manual assessment time by 40%.
Problem
Manual vulnerability assessment across large networks was time-consuming and inconsistent. Needed a scalable solution to identify and prioritize vulnerabilities across multiple endpoints.
Investigation
Analyzed existing scanning tools and methodologies. Identified bottlenecks in sequential scanning and false positive rates. Needed intelligent prioritization and comprehensive reporting.
Architecture
Custom scanner leveraging Nmap for network discovery and Metasploit for vulnerability validation. Integrated with lab environment for automated testing across 100+ users. Results aggregated and prioritized by severity.
Stack
Outcome
Reduced manual assessment time by 40%. Improved threat prioritization by 60%. Identified 150+ vulnerabilities across 50+ endpoints. Deployed for 100+ users in lab environment.
Implemented multi-modal biometric authentication system achieving 95% accuracy on 1,000+ samples. Reduced authentication time by 30% using Python and machine learning.
Problem
Traditional authentication methods were slow and prone to spoofing attacks. Needed a robust multi-modal biometric system for proactive defense research.
Investigation
Researched face recognition and voice authentication techniques. Studied attack vectors against biometric systems. Identified optimal feature extraction and fusion methods.
Architecture
Multi-modal biometric pipeline combining facial recognition and voice authentication. Real-time processing with ML-based liveness detection. Integrated anomaly detection for security.
Stack
Outcome
Achieved 95% accuracy on 1,000+ samples. Reduced authentication time by 30%. Contributed to proactive defense research initiatives.
Built Python CLI for data preprocessing with Pandas, NumPy, and Scikit-learn. Increased data cleaning speed by 30% and supported 10+ formats.
Problem
Data preprocessing was a bottleneck in data analysis workflows. Manual cleaning was error-prone and time-consuming. Needed a flexible tool supporting multiple data formats.
Investigation
Analyzed common data preprocessing tasks and formats. Identified performance bottlenecks in existing tools. Designed modular architecture for extensibility.
Architecture
Modular CLI tool with pluggable processors for different data formats. Leverages Pandas for efficient data manipulation and Scikit-learn for advanced preprocessing. Supports CSV, JSON, Excel, Parquet, and more.
Stack
Outcome
Increased data cleaning speed by 30%. Supported 10+ data formats. Adopted by 50+ academic users with 4.8/5 satisfaction rating.