CHART — PUBLICATIONS & PREPRINTS
Peer-reviewed work, preprints, and benchmark studies. The complete record below is drawn live from ORCID 0000-0002-0302-4812 — when the ORCID record changes, this page follows.
BENCHMARK STUDY
24 NIST post-quantum key-encapsulation mechanisms benchmarked head to head — timings, memory, and failure behavior, measured rather than quoted from spec sheets. Submitted to the 2025 AMA Research Challenge.
POSITION PAPER
A pragmatic standard for safe, secure, and trustworthy AI: drawn from clinical medicine's oldest rule and a red teamer's threat model, and written to be applied, not admired.
PEER-REVIEWED LETTER — CLINICAL ORTHOPAEDICS AND RELATED RESEARCH (2024)
On what generative AI actually changes in value-based care — where it earns its cost, where it does not, and the evidence bar it should be held to before it touches a patient.
PEER REVIEW
Ongoing peer review of manuscripts on LLMs in medical education and clinical AI — the same evidence standard, applied to other people's claims.
STAGE — DRAFTING · UPDATED 2026-10-12
A practice paper from building bunker: how hybrid classical/post-quantum narinfo signing works in a Nix binary cache, what the key-loading contract has to guarantee, and what a live deployment on qompass.ai proves out.
Complete the VPS deployment validation (push/pull as a signed substituter), then fold the results in.