Responsible AI
MedFlow uses AI to assist clinicians and hospital staff. It does not practise medicine, and it does not write to the patient record on its own.
Our principles
- AI-generated clinical content must be reviewed and approved by an authorised healthcare professional before becoming part of the patient record
- AI outputs are explainable in context: the clinician sees the draft alongside the record it was drawn from
- AI-assisted content is recorded as such, with the reviewing clinician and timestamp
- AI features are configurable and can be disabled per facility and per role
- Human accountability for clinical decisions always remains with the clinician
Where AI is used
- Drafting and structuring clinical notes and summaries
- Voice-to-structured-note support
- Decision-support prompts such as abnormal trends, unacknowledged results and risk flags
- Drafting assistance in radiology and laboratory reporting workflows
- Operational and analytical summaries for administrators
Limitations
- AI drafts can be incomplete or incorrect and must be verified
- AI features depend on an external model service and may be temporarily unavailable
- AI is not used to determine eligibility, triage outcomes or treatment without clinician confirmation
MedFlow home — Contact Ubora One Limited