Python-based cybersecurity platform integrating network scanning, traffic analysis, firewall log monitoring, phishing detection, and SIEM dashboards into a unified security monitoring solution.

• Automated network discovery and asset mapping
• Live packet capture and traffic analysis
• Firewall log monitoring and event correlation
• Phishing detection using heuristic analysis
• ELK Stack integration for centralized logging
• Docker-based deployment for easy setup

Built to unify reconnaissance, monitoring, and threat detection into a single security workflow, reducing manual effort while improving visibility into network security events.

AI-powered biometric authentication framework combining frequency-domain liveness detection, two-stage machine learning, and Honey Encryption to defend against face presentation attacks and securely protect biometric templates.

• Frequency-domain feature extraction using FFT and LBP
• Two-stage spoof detection with SVM/Random Forest and Autoencoder
• Confidence-based gray-zone verification for improved reliability
• Honey Encryption for secure biometric template storage
• Real-time webcam-based face authentication
• Modular architecture supporting training, evaluation, and deployment

Developed to strengthen biometric authentication systems by integrating computer vision, machine learning, and cryptographic protection into a unified anti-spoofing framework.

Implemented K-Means clustering to compress digital images while preserving visual quality.

• K-Means clustering for color quantization
• Image compression with reduced color palette
• Visual comparison between original and compressed images
• Adjustable number of clusters for compression control
• Machine learning–based image processing
• Lightweight Python implementation

Developed to demonstrate the application of unsupervised machine learning in image compression by reducing color redundancy while maintaining acceptable visual quality.

Machine learning model to classify SMS messages as spam or legitimate using the Naïve Bayes algorithm.

• Machine learning–based spam message classification
• TF-IDF feature extraction for text representation
• Multinomial Naive Bayes model for accurate spam detection
• Automatic preprocessing of email and SMS messages
• Interactive Tkinter-based desktop application
• Model evaluation using accuracy and confusion matrix
• Supports both training and real-time prediction
• Easy-to-extend architecture for future NLP enhancements

Biometric liveness verification system that authenticates users through randomized eye blink challenge–response sequences to prevent spoofing attacks.

• Randomized eye blink challenge–response authentication
• CNN-based eye state classification for accurate blink detection
• Real-time facial landmark tracking using MediaPipe
• CNN + Eye Aspect Ratio (EAR) fusion for robust liveness verification
• Active protection against photo and replay attacks
• Real-time webcam-based biometric authentication
• Modular architecture supporting training, testing, and deployment
• Configurable challenge sequences with confidence-based verification

Implements a challenge–response mechanism with randomized blink patterns to strengthen biometric authentication against presentation attacks.

Web-based employee management system for managing employee records and attendance.

• Employee registration and profile management
• Secure user authentication and authorization
• Attendance management module
• Administrative dashboard with reports
• SQLite database integration
• RESTful backend architecture using Flask
• Responsive frontend using HTML, CSS, and JavaScript
• Modular project structure for easy maintenance

Developed as a full-stack employee management solution that centralizes employee information, attendance tracking, and administrative workflows through a secure and user-friendly web interface.

Password security assessment tool that evaluates password strength using entropy calculation and security best practices.

• Password strength evaluation using entropy calculation
• Character set analysis (uppercase, lowercase, digits, symbols)
• Real-time password strength feedback
• Security improvement recommendations
• Random strong password generator
• User-friendly Tkinter graphical interface

Calculates password entropy and evaluates complexity based on length, character diversity, and randomness to encourage stronger password practices.