Review the suggested action
An authorized user considers the recommendation before the referral moves forward.
Human oversight
AI assistance. Human review and approval.
Elvra suggests what should happen next and presents the supporting evidence in an editable, source-linked draft. Authorized people verify the evidence and approve the action.
How human review works
Keep people in control at every important step.
Check the supporting evidence
The reviewer can return to the original pages behind the recommendation.
Refine
The recommendation and drafted reasoning remain editable throughout review.
Escalate
Uncertainty, exceptions and missing information can go to the right person.
Approve the action
Your organization defines who can approve each clinical or operational action.
Monitor the handoff
The approved action, ownership and activity remain visible after review.
Where review matters most
Give people a clear way to handle uncertainty.
- Poor-quality or unsupported documents
- Ambiguous, contradictory or incomplete information
- Specialty terminology and local routing rules
- Unusual cases and exceptions
- Information that requires confirmation against the source
Roles and responsibility
Match Elvra to your approval process.
Define who reviews a recommendation, who can approve the action and when work should be escalated. Elvra keeps the suggested action, reasoning, source, owner and status visible throughout that process.
Responsible-AI reference
What the NIST framework contributes.
The NIST AI Risk Management Framework Core recommends documenting human-oversight processes, responsibilities and accountability across an AI system’s lifecycle. Its human–AI interaction guidance also emphasizes clearly defining whether AI advises a human decision-maker or acts autonomously. These principles support Elvra’s review model; they are not evidence of clinical effectiveness.
See the control model inside the product.
Watch how draft review and source verification appear in the current walkthrough.