What information does Support need to troubleshoot my issue?

The information that our Support Engineers will find helpful in troubleshooting your issue will vary, but we have provided the following suggestions based on our products:

Batch API Transcription

• The request_id for the API request that you are troubleshooting

• The code that you are running to submit the request (if applicable)

• The audio that you are submitting with the request

• The full JSON response from our API (or any relevant HTTP status codes)

Streaming API Transcription

• The code that you are executing to initiate the and terminate the WebSockets stream

• The audio that is being read into the stream (if available)

• The WebSocket stream logs or API response for the behavior you are troubleshooting (as applicable)

ASR Model Accuracy

• The request_id for the transcription request from our API

• The audio that you are submitting with the request

• The full JSON response from our API (or any relevant HTTP status codes)

• The “truth transcripts” (human-verified transcripts) for the audio that was transcribed

Bear in mind that there may be more information needed to troubleshoot your issue, but these details should be a great starting point! 🙌