About

I am a security engineer and researcher. For the past decade, across DARPA-funded research and security engineering at Amazon and Meta, my work has pursued one question: how can the protection of sensitive personal data be made automatic, verifiable, and workable at real-world scale, not just on paper?

Today that question lives in two places: the large data systems that hold people’s most sensitive records, such as health data, and the AI systems and agents now entering regulated industries.

Now

I am a Security Engineer on Amazon’s AI & AGI Security team, where I work on securing foundation models. Before that, on Amazon Healthcare Security, I led the design of large-scale insider-threat detection for regulated healthcare data and threat-modeled AI-agent integrations built on the Model Context Protocol (MCP). Before Amazon, I was a Privacy Engineer on Meta’s Privacy Incident Response Engineering team, leading responses to high-severity privacy incidents and turning them into detections and preventive controls.

Before industry, I spent six years as a graduate researcher at the University of California, Irvine on the DARPA Brandeis TIPPERS project, building privacy-preserving IoT systems that were deployed on the UC Irvine campus and tested in the U.S. Navy’s Trident Warrior exercises.

Focus areas

  • AI and agent security: threat modeling for AI agents, indirect prompt injection, tool abuse, sensitive-data exfiltration, and security evaluation of foundation-model applications.
  • Detection engineering for sensitive data: insider-threat and data-misuse detection, validating detections before deployment, variant analysis, and incident-driven prevention.
  • Verifiable privacy: cryptographic attestation and privacy-preserving system design that make data handling provable rather than merely promised.

Highlights

  • 14 peer-reviewed publications in IEEE and ACM venues, 250+ citations (Google Scholar).
  • Best Paper Award, IEEE NCA 2019, for IoT Notary, tamper-evident attestation of sensor data.
  • DARPA-funded research in practice: co-developed the TIPPERS privacy platform, deployed on the UC Irvine campus and tested in the U.S. Navy’s Trident Warrior 2019 and 2020 exercises.
  • Program Committee, IEEE HealthSec 2026 (Workshop on Cybersecurity in Healthcare, co-located with ACSAC).
  • First place at the Linux Foundation Akraino 5G MEC Hackathon (2019) and the US Ignite SDN Innovation Challenge (2015).

Get in touch

I welcome invitations to review for security and privacy venues, to speak, and to collaborate on AI security and privacy engineering. The best way to reach me is email or LinkedIn.

Views expressed here are my own and do not represent my employer.