Skip to main content
Before you record any clinical audio, AfriHealth AI checks that your browser, microphone, and speech services are ready. This guide walks you through the System Setup & Hardware Diagnostics panel and then runs a sample case in the Frontline Triage module.

Prerequisites

  • A modern browser with microphone access (Chrome, Edge, or Firefox recommended).
  • The prototype served over HTTPS or localhost. Browsers block microphone capture on insecure pages.

Set up your system

1

Open the diagnostics panel

Open the prototype and select Diagnostics. The System Setup & Hardware Diagnostics panel opens.
2

Check the secure context

Browser Secure Context (HTTPS / Localhost) runs automatically. It must pass before the microphone can be used.
3

Test your microphone

Next to WebRTC Microphone Hardware Permission, select Test Mic and allow microphone access when your browser asks.
4

Ping the speech engine

Next to Intron Sahara API Endpoint Status, select Ping API to confirm the backend speech proxy is reachable.
5

Test voice output

Next to Speech Synthesis (AI Voice Response), select Test Voice to confirm your browser can read recommendations aloud.
6

Save your setup

Select Save & Return.
The EHR / FHIR Server Endpoint check is optional. The integration boundary is documented, but no production endpoint is configured in the prototype.

Run your first case

1

Pick a Gold Standard case

In Module 1: Frontline Triage, select Case 1: Acute Emergency (Chest Pain & Fever).
2

Process the audio

Select Pre-load Audio to use the sample recording. Use Judge Mode (No Mic Req.) if you have no microphone.
3

Review the output

Check the transcript, Extracted Medical Entities, and Triage Classification.

Result

You see a code-switched transcript with a confidence score, extracted entities, and a draft referral recommendation for clinician review.

Next steps