Creating a New EMR

Doctor

This is Medisa's main flow: record the consultation, then get back a SOAP draft ready for your review.

Before you start

Signed in, with a workspace selected. See Authentication.
Microphone permission granted in the browser.
Recording quota available — the remaining balance is shown on the home screen.
The patient already registered, or ready to be added on the spot. See Creating a Patient.

Steps

On the home screen, press Start Recording.

Medisa home screen with the start recording button, an upload audio button, plan summary, and remaining quota — The button to its right uploads an audio file you already have.

Confirm the recording consent. By continuing, you state that the patient has agreed to being recorded.

Recording consent and data protection dialog with cancel and continue buttons — Do not refresh or close the page while recording runs.

Choose whether this is a registered patient or a new patient.

Patient type dialog offering registered patient or new patient

Find and select the patient from the list.

Select patient dialog listing patients with their record numbers and phone numbers — Search by name or medical record number.

The selected patient appears above the recording panel. Press Start Recording again to actually begin.

Home screen with a banner reading Patient: Patient 1 and a Change link on the right — Wrong patient? Press Change before you start recording.

Hold the consultation as usual. While recording, audio is saved automatically every 30 seconds to your device.

Home screen while recording, showing a live waveform and a red Stop & Process button — While the Stop & Process button is visible, recording is running.

When you are done, press Stop & Process. The recording uploads, and Medisa transcribes it and composes the SOAP note. Progress is shown until it finishes — you can leave the page and come back later.

Processed recordings appear in the Medical Records list on the home screen.

Medical records list on the home screen showing patient name, EMR number, date, and verification status — An Unverified status means the draft is waiting to be reviewed.

Open the record to read the result, then press Verify once the content is correct.

Medical record detail with audio player, EMR and Transcript tabs, cited SOAP content, an ICD-10 code, and Edit EMR and Verify buttons — The Transcript tab shows the original conversation text.

Reading the result

An AI draft always needs a doctor's review

The generated SOAP note is a draft, not a final document. While it is still Unverified, you have not signed off on it. Read it, correct what needs correcting, then verify.

A few things that help while reviewing:

Citations such as s4 or s9 mark the part of the transcript a point came from. Use them to check quickly whether the AI heard it right.
ICD-10 codes in the Assessment section are suggestions and still need your verification.
The audio player at the top lets you replay anything you doubt. See Transcription & Audio.
Need to change the content? Press Edit EMR. See Editing an EMR.

Language

The transcript follows the spoken language

The spoken language is detected automatically, and the transcript is always in the language of the conversation — never translated or overwritten. The record detail shows the original language, for example "Original language: Bahasa Indonesia".

Only the SOAP content can be translated, and that happens in the background without holding up saving. Any translated view is labelled as AI-generated.

If the connection drops

Recordings are written to device storage first, not straight to the server. So:

Losing connection mid consultation does not lose the recording. The upload resumes on its own once the device is back online.
App closed unexpectedly? The last automatically saved segment is still on the device.
Failed uploads are retried a few times. If they still fail, the recording is marked as failed and you can resend it manually.
Resending the same recording does not consume quota twice.

Uploading existing audio

If the consultation was recorded with another device, use the upload button to the right of the record button. Supported formats are MP3, WAV, M4A, WEBM, and OGG, up to 50 MB per file. Everything after that is identical to a live recording.

If something goes wrong

Symptom	Most common cause
The record button does nothing	Microphone permission has not been granted in the browser
Recording will not start	Quota exhausted — check the remaining balance on the home screen
A recording sits in processing for a long time	The connection dropped; leave the app open once back online
The SOAP note feels thin	The audio was unclear — microphone too far away or a noisy room
The recording went to the wrong clinic	The active workspace was not the one you intended
Wrong patient on the record	The patient is bound before recording starts; use Change next time

Creating a New EMR

On this page