Skip to content

arulagrawal/i_hate_voice_notes

Repository files navigation

I hate voice notes

A Whatsapp bot which replies with text transcripts of audio you send it.

The intended workflow is to forward voice notes your "friends" send you to the bot, and it will reply with the text transcript.

Setup

You need to set 3 environment variables.

WHATSAPP_TOKEN your token for your app from Meta.

VERIFY_TOKEN your custom verify token that you give Meta.

OPENAI_API_KEY your api key for openai in order to use Whisper.

Problems

  • The bot gives you no indication it's working until you either get or don't get a transcript.
  • The code downloads the audio to a file then uploads that file again. It should be possible to do this in memory.
  • create a docker image

About

WhatsApp bot that transcribes voice notes.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published