You send a voice sample to Telegram — the scenario makes a digital copy of it, voices the needed text with it and saves the recording to Google Drive.