Whisper service

Transcribe audio via API.

Demo: https://whisper-app-b4lxkp5rjq-uc.a.run.app/api/audio/?url=https://eslyes.com/easydialogs/audio/dailylife013.mp3

API

GET Get audio transcription

GET /api/audio?url=<audio_url>

POST Schedule audio for transcription

/api/audio
{
  url: audio url,
  ADMIN_ACCESS_TOKEN: access key to schedule audio
}

Design

Service consists of app and worker:

• App exposes endpoints to return and schedule audio transcription
• Worker transcribes the audio and saves transcription to MongoDB