Pilot Program
Welcome to the Apodixis Pilot
Independent verification for AI-generated clinical documentation
What Apodixis Does
Extracts Claims
Reads the clinical note and identifies every medication, diagnosis, vital, lab, and allergy claim.
Verifies Against EHR
Cross-references each claim against the patient’s FHIR record for accuracy.
Flags Discrepancies
Highlights contradictions, wrong doses, hallucinated medications, and omissions.
How to Read a Verification Report
Verification Score
Each note receives a score from 0 to 100. Higher is better. A score below 70 indicates significant issues were found and the note warrants careful review before signing.
Claim Statuses
Severity Levels
DAX vs Manual Notes
During this pilot, every note — whether generated by DAX or written manually — is verified. This lets us measure the accuracy difference and identify where AI scribes need improvement.
Giving Feedback
If you see a flag that is wrong, mark it as “Dismiss” with a reason. This helps us tune the system. Your feedback directly improves accuracy.
Frequently Asked Questions
Does this slow down my workflow?
No. Verification runs in shadow mode. You review results after signing. There is no interruption to your current charting process.
Is my data safe?
Yes. We never store raw note text. Only hashed references are retained in our database. Clinical note text is processed ephemerally and automatically deleted within minutes. Full HIPAA compliance with a signed Business Associate Agreement.
What if the system flags something incorrectly?
Dismiss it with a reason. We track false positives and continuously improve the system based on clinician feedback. Every dismissal makes the next verification more accurate.
Who sees my verification results?
Only you and your department admin. Results are private to your organization and are never shared externally.
Questions about the pilot? Contact your department administrator or reach us at chris@apodixis.ai.