STEP-BY-STEPvoice.aiconcept.art≈ 9 min
From recording to transcript in Voice
Record or import a WAV, transcribe, check it by replaying the matching range, then export.
What you end up with
You end with a transcript you have checked by ear against its own audio range, and an exported file.
Cost
We have not measured the cost of this flow in the interface, so no figure is quoted. Where the product prints a price itself, we quote that exact number.
Before you start
- A signed-in AIconcept account
- A microphone, or a WAV file to import
Steps
Step 1: Open the record tab
The application is organised in tabs; start on the record tab.
Label on screen
Record · Ghi âmStep 2: Record live, or import an existing WAV
If you already have a meeting recording, importing the WAV is faster than re-recording. Stop and cancel controls are available mid-run.
Label on screen
Import WAV · Stop/cancelStep 3: Run the transcription
The application announces progress through aria-live regions, so a screen reader speaks state changes without you hunting for them.
Label on screen
TranscriptionStep 4: Check it by replaying the matching range
Range playback lets you hear the exact audio behind a doubtful line instead of scrubbing by hand.
Label on screen
Play rangeStep 5: Export the result
The export tab is the last step; check the transcript first, because fixing it after export means doing the work twice.
Label on screen
Export
What this lesson does not promise
- This application's interface is currently mostly English while the rest of the platform is Vietnamese. It is a recorded inconsistency being fixed; this tutorial keeps the English labels because that is what is on screen today.
- The AICoin cost of transcription has not been measured in the interface, so no figure is quoted.
How this lesson was verified
Measured while signed in: 47 controls, 23 inputs, and 11 aria-live regions — the most of all 12 applications.
Source: docs/audit/current/2026-08-13-authenticated-ui-inventory.md