Hello together, I want to achieve the following: Slack Message with keyword and voice message → send to Whisper → transcribe voice message.Process:Every morning the Slackbot sends a message to my team asking them for a quick checkin on what they plan for the day. They answer to this Slackbot with a message containing only “dailycheckin” + the voice message. I then want to transcribe this voice message. It works perfect, when I send messages to the Slackbot. It does not work, when my team responds to these messages in their Slackbot.Do I need to use a public channel? Because I dont want to do this because of confidential voice messages.I use Trigger: “New Message from Query” - Search Query: dailycheckinRegards,Soeren

Best answer

How do I transcribe Slack Voice Message with Whisper for whole Team

Hello together,

I want to achieve the following: Slack Message with keyword and voice message → send to Whisper → transcribe voice message.

Process:

Every morning the Slackbot sends a message to my team asking them for a quick checkin on what they plan for the day.
They answer to this Slackbot with a message containing only “dailycheckin” + the voice message.
I then want to transcribe this voice message.
It works perfect, when I send messages to the Slackbot.
It does not work, when my team responds to these messages in their Slackbot.

Do I need to use a public channel? Because I dont want to do this because of confidential voice messages.

I use Trigger: “New Message from Query” - Search Query: dailycheckin

Regards,

Soeren

Best answer by pranayshindeBest answer by pranayshinde

When your team responds with a voice message, Slack provides a private URL for the audio file.

You’ll have to pass the extracted URL to the Whisper API somehow and whisper also has its limitations, but a workaround is provided here:
https://community.openai.com/t/whisper-can-we-transcribe-from-url-and-file-upload/95700/14

Did this topic help you find an answer to your question?

+5

When your team responds with a voice message, Slack provides a private URL for the audio file.

You’ll have to pass the extracted URL to the Whisper API somehow and whisper also has its limitations, but a workaround is provided here:
https://community.openai.com/t/whisper-can-we-transcribe-from-url-and-file-upload/95700/14

Rich Text Editor, editor1