AISafe Job API – Documentation

Note: These APIs are available for private deployment. Not Generally Available. Email our team for getting access to these APIs: info@mayadataprivacy.eu

1. Submit a Transcription Job

Endpoint:

POST /mdp/ai-safe/job/transcribe

Authentication:

• Type: API Key
• Key Name: Authorization
• Location: Header
• Value: Bearer <JWT TOKEN>
• Description: JWT token for authorization. You can obtain your API key from User Authentication API.

Description

Submits a new transcription job with an audio file. The file is processed asynchronously. Once completed, the result will be posted to the given webhook URL.

Request

Content-Type: multipart/form-data

Name	Type	Required	Description
file	File	✅	Audio file to transcribe (.mp3, .wav, etc.)
lang	String	✅	Language code (en, de, etc.)
webhook_url	String	✅	HTTPS endpoint to receive the result

Example Request (cURL)

curl -X POST https://yourdomain.com/mdp/ai-safe/job/transcribe \
  -H "Authorization: Bearer <JWT TOKEN>" \
  -F "file=@/path/to/audio.mp3" \
  -F "lang=en" \
  -F "webhook_url=https://client.com/transcription-webhook"

Response

{
  "status": "accepted",
  "job_id": "c751c050-1f7d-424b-a0a4-d5f4cc8bd682"
}

---

2. Webhook Payload

Once processing completes, the API sends a POST request to the provided webhook URL.

Success

{
  "job_id": "9e64b8a2-4d15-4562-a68b-df39efb6e771",
  "status": "completed",
  "transcript": "Hello, this is a transcribed audio file."
}

Failure

{
  "job_id": "9e64b8a2-4d15-4562-a68b-df39efb6e771",
  "status": "failed",
  "error": "Audio format not supported"
}

---

Supported Audio Formats

• .mp3
• .wav
• .m4a
• .ogg

---

Notes

• The webhook endpoint should be accessible from the AISafe deployment environment and must not have internal-only network restrictions.
• Webhooks are sent with Content-Type: application/json.
• Jobs are queued and processed asynchronously.
• If processing fails, a "failed" status will be sent with an error message.