Through the whole visit the doctor was typing, eyes on the record instead of the person in front of them, and then finishing the rest of the notes at home after dinner. The documentation was eating the care, and the clinicians were burning out doing it.
Every visit was a split screen: listen to the patient, or type into the record. The doctor could not fully do both, and the record usually won, because an unfinished note follows you home.
And it did follow them home. Clinicians routinely finished their documentation after hours, the so-called pyjama time, which is exactly the kind of unpaid, invisible overload that drives good doctors to burn out or cut their days.
It also capped throughput and quality. Time spent typing was time not spent seeing patients, and notes written late or from memory were thinner and slower to reach the record, which matters for the next clinician who reads them.
Listen in the room, draft a structured note into the record, and keep the clinician in control of every word that is saved.
The engine transcribes the doctor-patient conversation in the room in real time, so the clinician can look at the patient and talk to them instead of narrating to a keyboard.
It drafts a structured visit note, complaint, findings, assessment and plan, in the record's own format, so what lands is a usable clinical note rather than a raw wall of dialogue someone still has to rewrite.
The follow-ups mentioned in the visit, referrals, prescriptions, tests to order, are picked up automatically and proposed in the record, so the small things that used to slip when notes were written later are caught in the moment.
The draft surfaces for the doctor to edit and sign before anything enters the record, so the clinician stays fully responsible for the note and the AI never writes the chart on its own.
The note is drafted straight into the electronic health record the clinic already uses, so there is no second system to visit and no copy-paste between tools at the end of a long day.
With the typing lifted, the encounter becomes a conversation again, which is both a better experience for the patient and the reason clinicians stop dreading their own documentation.
Less time typing, no after-hours backlog, and doctors back with their patients.
The time each doctor spent on documentation fell sharply, and the after-hours note-writing that used to eat their evenings largely disappeared, because the note is drafted during the visit rather than reconstructed from memory afterwards.
Freed from the keyboard, clinicians spend the visit facing the patient again, which lifted both the experience in the room and the clinicians' own load, with every note still reviewed and signed by the doctor.
A medical record is not a place for an unchecked machine, so the discipline is the point: capture the conversation, draft a real note, and put the doctor's review and signature between the draft and the chart. It is the same draft-from-the-conversation, human-signs-off discipline behind our after-call-work work, aimed at the exam room rather than the call centre.
A provider's prior-auth requests took days and many bounced back denied on a technicality, writing off revenue. AI builds each request to the payer's own rules, flags what's missing, and drafts appeals for a manager to approve.
Read the case → Reputation management · HealthcareA clinic network's Google rating sank because only unhappy patients reviewed. An automated feedback loop routed happy patients to public reviews and unhappy ones to a manager to call, lifting the score from 3.8 to 4.8 and doubling organic leads.
Read the case →
Reef TRH