Mohamed Amgad Khater
Security R&D Tech Lead
I’m Security R&D Tech Lead at Cyshield, where I build AI-native security products and the model infrastructure they run on: an application security platform, agentic SOC systems in daily use by our analyst teams, and the security architecture behind a commercially launched AI-powered WAF.
Most of that work rests on the model layer underneath it, so a fair amount of my time goes into AI hosting and evaluation: deploying and tuning open-weight models on vLLM and llama.cpp, building the harnesses that let one security core run against a local or a hosted model, and running the benchmarking work that decides whether an AI-assisted approach actually finds more than what it replaced.
Before I built security products I broke them, and the offensive years are still what makes the building work. I hold OSAI (OffSec Advanced AI Red Teaming), OSWE, eWPTXv2, eCPPTv2 and eMAPT, alongside CVE research (CVE-2017-1000058, CVE-2018-5222) and disclosures to Fortune 500 companies through HackerOne and Bugcrowd. I’m also the creator of DVBLab, a banking security training platform. I write rules against the bypasses I would use myself, and I test agent systems the way I would test the ones I ship.
Interested in collaboration, security research discussions, or consulting opportunities? Connect via LinkedIn, explore my GitHub, view my complete resume, check my certifications, or read my latest research.