The intelligence layer that sits on top of the EHR. It reconstructs patient journeys, finds the administrative delay inside them, audits documentation, flags duplicate and inconsistent records, and writes an executive summary. It is also wrong often enough that the only safe way to use it is the one this laboratory drills: investigate, then use the AI, then verify it, then decide.
Eight cases. Every one runs the same loop — open the EHR evidence, run the AI analysis, verify the AI output against the record, take an administrative decision, and record the action. The verify step is not decoration: in three of the eight cases the AI states something the record does not support, and one of those would enter a patient's permanent history if you accepted it.
Two indices sit alongside your mark. Verification discipline falls when you decide without opening the evidence. Governance integrity falls when an unverified AI statement is accepted, patient data is handled carelessly, or an irreversible action is taken on a probability.
Resolve the case library to generate your report.