Feature request: Email notifications when transcription is complete

[devkit415]

Hello!

I’d like to set up an email notification for users whenever a recording’s transcription has finished successfully or encountered an error.

According to the documentation, webhooks can listen for this event:

recording.transcription.completed

I also understand that each webhook belongs to the user who created it. However, the webhook sends the notification to an external URL, and the payload doesn’t seem to include the user’s name or email address.

Is there currently a built-in way for Speakr to automatically email the user who owns the webhook when an event such as recording.transcription.completed is triggered?

In particular:

• Can Speakr send emails directly for events such as recording.transcription.completed?
• Could the webhook payload include the webhook owner’s name and email address?
• Is there a variable, template, or admin setting that could enable this?
• Is the expected approach to receive the webhook through an external service, look up the user by user_id, and then send the email via SMTP?
• If this isn’t supported at the moment, would it be possible to add it as an optional feature or extension?

Thanks in advance, and congratulations on your great work!

[murtaza-nasir]

Your reading of the code was right on both counts, and it is now built.

v0.10.6-alpha adds an opt-in setting under Account, Preferences: "Email me when a transcription finishes". It sends a message as each recording completes and again if one fails permanently, using the same SMTP configuration as email verification and password resets. Set APP_BASE_URL to your instance's address so the message can link back to the recording.

The webhook payload still carries user_id and not the address, deliberately, since a webhook belongs to the user who created it and only fires for their recordings, so the receiver already knows whose events are arriving. Two things you should know if you are integrating against it: audio_duration_seconds and language were documented but never sent, and both are sent now. The language is the one the transcription service detected for that recording, so it will be absent for recordings transcribed before this release, and for backends that report none.

Summaries do not send mail, since a summary follows a transcript you have already been told about. If a per-instance default would be useful instead of per-user opt-in, say so and I will look at it.

[devkit415]

Thank you very much for the quick response, the clarifications, and for implementing this feature!

I'll try out the new version as soon as I can.