Documentation navigation
Safety & safeguards
Menaela is designed to act inside explicit clinic data, permissions and workflow boundaries. When those are not enough, it should clarify or involve staff.
In this pageThe main control layers · Facts and actions come from the system · Clinical judgment remains a boundary
The main control layers
- Clinic autonomy switch for automatic external replies.
- Visible autonomy status for staff.
- Staff-attention records for questions that need manual or clinical review.
- Human takeover for a specific patient conversation.
- Permission checks based on the signed-in staff user.
- Approval controls for sensitive actions.
- Clinician review and signing for AI Scribe drafts.
Facts and actions come from the system
Menaela should retrieve availability, providers, services, prices and patient data from the relevant clinic records rather than generating them from model memory. Before sending an autonomous message, the system also rechecks whether the conversation is still eligible for AI delivery.
Clinical judgment remains a boundary
Patient-facing Menaela can answer supported clinic information and administrative questions. Personalized medical judgment requires staff review rather than a confident autonomous answer.