At Vontier, I built an AI-powered threat modeling tool that replaced Microsoft TMT and cut the time for 30 threat models from one month to one week. I also co-led STRIDE and attack-tree workshops that surfaced 300+ design flaws before release across IoT, EV charging, and cloud products.
In my current role as a Junior Product Security Engineer, I harden products through secure code reviews and penetration tests, and work with developers on fixes. I tuned GitHub Advanced Security scans across 300 repositories, improving detection accuracy by 30% and cutting false positives.
My projects include theriac, an open-source scanner for hidden instructions in MCP tool descriptions, and ADAM, an LLM agent for autonomous lab-target penetration testing. I also founded Sealed, a service that finds and fixes security holes in AI-built apps.

