Ambience Prepare Surfaces Evidence and Open Questions Before Visits

Ambience Healthcare on November 12, 2026 introduced Ambience Prepare, a pre-visit workstream whose Chart Prep feature turns longitudinal patient records into evidence-linked briefings. The system is designed to investigate the record around an upcoming encounter, highlight unresolved questions and leave clinical decisions with the clinician.

Prepare also includes Chart Chat, which lets clinicians ask natural-language questions about a patient’s chart. The announcement focuses on Chart Prep, built for the short review window before a visit: orient the clinician, restore the relevant care thread, highlight changes and surface open questions. Supporting evidence remains attached to findings, and incomplete or conflicting records are flagged.

Connecting Medication Records Before Surgery

Ambience illustrates the distinction with a urologist preparing to see a patient with a recent multifocal stroke before a planned transurethral resection of the prostate. Cardiology had planned to change medications if atrial fibrillation was detected. A later test confirmed atrial fibrillation, but subsequent records left unclear whether apixaban was started and clopidogrel stopped.

Each entry could be summarized accurately without resolving the medication question central to surgical preparation. Chart Prep is instead intended to connect the procedure with medications relevant to bleeding risk, follow the cardiology plan, check the rhythm-monitor result and compare the intended change with the later medication reconciliation. It shows what the record supports, identifies conflicts and preserves unresolved questions for clinician review.

Chorus Supplies Shared Patient Context

Chart Prep draws on Chorus, the context system Ambience announced on August 24, 2026. Chorus gives the agent tools to survey a record, follow a clinical thread, open underlying evidence, compare sources and trace conclusions to their origins. It maintains a shared, source-linked understanding of the patient that different Ambience workflows can reuse.

Ambience argues that loading as much chart text as possible into a model is insufficient: long records can exceed context windows, crowded context can be harder to use reliably, and the model still needs direction about which threads and evidence to inspect. The earlier Chorus announcement describes navigation across notes, medications, labs, admissions and diagnostic reports, with reconciliation of conflicting sources and timelines. The electronic health record remains authoritative, while Chorus supplies a shared context layer between that record and Ambience’s AI workflows.

Evaluation Tracks Clinically Important Facts

Chart Prep’s evaluation starts with board-certified clinicians reviewing de-identified charts as they existed immediately before an upcoming visit. They identify and rank encounter-relevant facts; a second clinician independently reviews the case, with disagreements or omissions adjudicated. Automated reviewers compare outputs with those clinician-established references only after reaching acceptable agreement with clinician judgments. Each case is run multiple times to assess consistency.

On a benchmark of 46 de-identified patient cases with three independent runs per case, Ambience reports that its newest context harness, which shapes how the agent investigates a chart, surfaced 92.2% of critical facts, compared with 78.7% for the production harness. The 95% confidence interval for the newer result was 87.2% to 96.3%. Recall of all critical and high-importance facts rose from 66.3% to 72.3%.

The share of remaining misses attributed to failure to retrieve necessary evidence fell from 19.9% to 9.3%, a reduction Ambience reports as 53%. The comparison held the generation model, patient cohort, clinician references and evaluation suite constant, leaving the Chorus-built context harness as the principal change. The measured difference therefore concerns how the agent investigates the chart rather than a change in the underlying model.

That distinction gives health systems a specific measure of the product’s usefulness: whether better investigation retrieves and prioritizes facts clinicians consider consequential while preserving evidence and uncertainty for review. In the medication example, useful preparation makes the unresolved plan visible before surgery; deciding how to reconcile it remains with the clinician.