The 30-Day Pilot

A four-week supervised pilot scoped to one site or specialty, on real production volume. Radiologist signs every report. AI is the assistant, never the author.

Four Weeks, Four Modes

Week 1

Setup & Baseline

Integration verification, baseline metrics capture, radiologist orientation.

  • PACS / HL7 / ORU connectivity check
  • Voice-profile & macro audit
  • Baseline TAT, edit-distance, sign-rate
  • Per-radiologist orientation session
Week 2

Shadow Mode

AI drafts every case in parallel. Radiologists report normally. Drafts are visible but unused.

  • Side-by-side comparison logging
  • Critical-finding flag verification
  • Daily ops check-in
  • No production signoff changes
Week 3

Supervised Co-Pilot

Radiologists use AI draft as the starting point. Every edit is tracked. Signoff remains fully radiologist-controlled.

  • Edit-distance per radiologist per study
  • Sign-time delta vs. baseline
  • Anonymous radiologist feedback
  • Critical-results pathway retest
Week 4

Evaluation & Decision

Quantitative evaluation, qualitative debrief, and a written go / no-go recommendation.

  • Full metrics report delivered
  • Radiologist NPS survey
  • Site-specific deployment plan
  • Optional: extended pilot for additional sites

What You Get Out of the Pilot

Quantitative Deliverables

  • TAT comparison: baseline vs. pilot (study-available-to-final-signed)
  • AI-suggestion-to-radiologist-accepted time
  • Edit-distance metrics per radiologist and per study type
  • Sign-rate and addendum-rate comparison
  • Critical-results pathway compliance verification
  • Audit-trail integrity verification (ORU^R01 round-trip)

Qualitative Deliverables

  • Anonymous radiologist NPS and written feedback
  • Workflow friction log (with site-specific resolutions)
  • Voice-profile and macro migration plan
  • Integration risk register
  • Recommendation: extend / broaden / hold

Pilot Exit Criteria

TAT improvement (study-to-signed)
≥ 25% reduction vs. site baseline, sustained over the second half of Week 3
Edit-distance threshold
Median radiologist edit-distance below the agreed per-study-type threshold; no outlier study type above 2× the median
Critical-results compliance
100% of critical findings surfaced by the model are flagged in the radiologist workflow; zero missed escalations in pilot audit
Radiologist NPS
Group-wide NPS neutral or positive; no individual radiologist veto
Audit trail & ORU integrity
Every pilot-signed report reaches the RIS with intact edit trail and radiologist attribution
Operational stability
≥ 99.5% uptime during Weeks 2–4; no P0 incidents unresolved within 24h

Scope a Pilot With Us

Send us your site details — modality mix, current TAT baseline, approximate monthly volume — and we'll come back within two business days with a tailored pilot plan and integration checklist.

Request a Pilot Plan