AI infrastructure security / agentic threats / verification
Offensive security depth for frontier AI systems.
I find where complex systems can fail, build technical ways to prove the risk, and turn that evidence into concrete remediation, policy, and organizational change.
10 years security depth
Full-stack penetration testing, security engineering, red teaming, within high-stakes environments.
AI security practice
Guiding generative and agentic AI security work while building independent frontier AI security research.
Vulnerability research
Selected findings include a Bluetooth library supply-chain takeover vulnerability affecting 350M+ live devices and a mobile device lockscreen bypass affecting billions of devices.
Research track
Heron AI Security Fellowship Research Lead, AGI strategy training, biosecurity training, and a RosettaCon talk on risks at the intersection of LLMs and biodesign tools.
Turn sharp threat models into measurable research directions.
Review a system end-to-end, mapping out assumptions, weak points, and critical assets. Separate signal from noise, and create hypotheses to test.
02 / Failure modeProve the technical risk early enough to fix.
Use offensive security instincts to locate failure modes, gather evidence, and design their control points before risk becomes catastrophic.
03 / ActionBuild the bridge from risk to action.
Translate technical results across audiences to maximize global impact.
Measurable and actionable
Security research should turn invisible risk into work people can act on.
My focus is technical evidence that can move a system, a team, or a decision: not just a warning, but a measurement with a next step.