HIPAA-by-design
Audit-log retention. Role-scoped access. No PHI in logs, ever. Phase 2a operates against the smallest surface, not the patient surface, and the architecture enforces that boundary.
Snap a prefcard with your phone, scan a barcode, or stream from a wearable. AI vision cross-references every item against the FDA GUDID, so the card on file matches the back-table on the day. No more stale cards.
Three movements. Phone by default, barcode when it is faster, wearable streams when a pilot service line wants hands-free.
Photograph the prefcard with your phone, scan a barcode, or stream from a wearable. The OR team stays heads-up; the capture surface does the looking.
An AI pipeline cross-references each scanned item against the FDA GUDID and manufacturer catalogs in real time. Every match traces back to an authoritative device record, never a guess.
Surgeon-facing preference cards stay accurate, auditable, and up-to-date automatically. Discrepancies surface in a review queue; nothing changes on the card without a human approving it.
Pull back the curtain on a working surface.
No fake screenshots. The link below opens the parchment-cream picker we ship to OR staff and surgeons today. You can browse, open a card, and see the supply list and provenance footer. Honest empty when data is not there yet.
Audit-log retention. Role-scoped access. No PHI in logs, ever. Phase 2a operates against the smallest surface, not the patient surface, and the architecture enforces that boundary.
Every match traces to an authoritative device record. If GUDID cannot confirm an item, the card flags it for human review. We would rather show you a gap than fabricate confidence.
Phone is the default. Wearables are pilot tier: Iristick G2 PRO and OpenMV RT1062 for service lines that want hands-free. Ray-Ban Meta in demo.
We start narrow: one service line, one room, one set of cards, and let the GUDID matches build the case. You will know inside the first month whether the drift is real. We will not ask for a logo on this page until it is.