Inside the Unified Wound Record: One Data Model for Six AI Agents
product
A technical look at the single wound record that powers WoundScribe's six agents — schema, EMR writeback, and why it eliminates duplicate charting.
The reason six agents can co-care without stepping on each other is a single wound record. Not six copies synced over API — one record, one source of truth.
The schema
Each wound has a persistent ID that carries:
- Images and 3D measurements captured over time
- Tissue composition, exudate, undermining, tunneling
- SOAP note history generated by the scribe agent
- ICD-10, CPT, HCPCS codes with modifier logic
- Healing trajectory and stall flags
- Patient education sent and acknowledged
Why one record matters
In a patchwork stack, the imaging tool stores measurements in its own database, the scribe writes a PDF to the EMR, and the coding tool reads a different field. When they disagree — and they do — the clinician reconciles by hand. One record removes the reconciliation step entirely.
How agents read and write
Each agent has scoped permissions on the record. The imaging agent writes measurements. The scribe agent reads measurements and writes the note. The coding agent reads the note and writes codes. The healing dashboard reads everything and writes trajectory flags.
EMR writeback
The unified record pushes structured data back to your EMR — Epic, Athena, eClinicalWorks, or the AI-powered EMR for wound care if you use ours. No PDFs, no manual re-entry.
What this eliminates
Duplicate charting, measurement drift between tools, code-to-note mismatches at audit, and the clinician tax of logging into five apps per encounter. For related architecture reading, see the not-so-hidden cost of patchwork tools.