Projects
Malware Detection Framework
A real‑time system leveraging deep adversarial networks and graph‐based feature extraction to accurately identify and quarantine emerging malware threats. This system utilizes transfer learning and ensemble methods to improve detection accuracy and minimize false positives.
Learn More
Prompt Injection Attack Mitigation
An investigative system that simulates prompt-based attacks on large language models and integrates real-time filtering algorithms. Uses adversarial training to reduce injection vulnerabilities while preserving the natural language processing capabilities.
Learn More
Graph-Based Intrusion Detection
Utilizes graph convolutional networks to analyze complex network traffic patterns, modeling endpoints as nodes and interactions as edges. This approach helps detect subtle structural anomalies indicative of cyber intrusions.
Learn More
Anomaly Detection Dashboard
A dynamic dashboard that aggregates data from multiple security sensors to visually highlight anomalous patterns. Leverages clustering algorithms and statistical analyses to enhance real-time situational awareness.
Learn More
AI Security Sandbox
An experimental platform to simulate and stress-test AI security protocols under controlled attack scenarios. Employs synthetic data generation and adversarial benchmarking to assess model resilience.
Learn More
AI-Driven Data Visualization
A robust platform that converts complex cybersecurity datasets into interactive visual insights. Utilizes machine learning to identify critical trends and integrates modern D3.js-based visualizations for intuitive analysis.
Learn More