Measure value in your own workflow.
Free early access gives your team a way to evaluate report QA under clinician supervision. We do not yet publish validated customer outcomes or independently measured accuracy figures.
Start with something you can try.
Explore a labelled synthetic case, inspect the workflow screens and review the data-handling explanation. These demonstrate the product; they are not customer testimonials or proof of clinical performance.
What the workspace dashboard measures
Signed-in workspace results use recorded activity and a selected date range. No fabricated success percentages or customer logos are used.
| Measure | Recorded evidence | Interpretation |
|---|---|---|
| Alert usefulness | QA runs, alerts, rated alerts and useful / not-useful clinician ratings. | Show rating coverage and counts. Unrated alerts are not useful ratings; multiple reviewers may rate an alert. This is not sensitivity or specificity. |
| Editing activity | Draft-save events, distinct edited reports and signed reports marked edited. | Activity is a proxy for effort, not a count of corrected errors or a measure of typing time. |
| Reporting elapsed time | Median and 90th-percentile time from review creation to signature, for eligible signed reports. | Includes waiting and interruptions. It is not active clinician time or measured time saved. |
| Adoption | Distinct participants with tracked workflow events during the chosen period. | Participation does not establish retention, clinical benefit or a share of all eligible users. |
A practical evaluation plan
-
Agree the scope
Choose report types, supported inputs, reviewers, usage capacity and data-handling arrangements.
-
Define a baseline
Record a comparable review workflow before evaluation. Document dates, sample size, exclusions and case mix.
-
Review both flags and misses
Clinicians inspect source evidence, rate alerts, correct reports and separately assess missed issues. Include reports without alerts.
-
Report counts and limitations
Share denominators, rating coverage, elapsed-time definitions and unresolved issues. Publish a customer case study only with permission and measured evidence.