Pathik Gohi
Biography
Certified penetration tester and offensive security professional with hands-on experience in red team operations, full-stack web application testing, and enterprise threat simulation. Proven expertise in secure architecture design, vulnerability research, and security consulting for global clients. Skilled with frameworks like MITRE ATT&CK and tools including Cobalt Strike, Burp Suite, and BloodHound. Currently pursuing an M.Eng in Cybersecurity at Hosei University
Research Interest
Al-Driven Runtime Defense Against Zero-Day Exploits
Abstract
Zero-day attacks and delayed patch deployment continue to expose critical systems to active exploitation. This presentation introduces an AI-driven patchless protection framework that delivers real-time, autonomous defense without requiring source-level code changes or system downtime. The approach combines dynamic memory instrumentation, behavioral modeling, and algorithmic decision intelligence to detect and mitigate exploitation attempts during execution. Machine-learning models such as Random Forest and Isolation Forest analyze live telemetry to identify anomalous activity, while adaptive feedback logic refines defensive responses under evolving threat conditions. Experimental demonstrations highlight significant reductions in attack success rates and improved operational resilience in unpatched environments. By integrating AI reasoning, virtual patching, and self-healing runtime control, this work showcases how organizations can transition from reactive patch cycles to proactive, adaptive, and sustainable cybersecurity across enterprise, IoT, and cloud infrastructures
