Inspiration

Focusing on a simple use case for a Poe bot, motivated by doing a single thing well.

What it does

Provide an audio file (or link to URL), get a transcript!

How we built it

Built using Poe's bot API for the bot interface, and fal.ai's Whisper inference model for the transcription.

What's next

There are a few things I didn't get to but see as good extension opportunities:

  • Adding limits to the provided audio file to keep costs down and the user experience good
  • For longer audio, offering an LLM-powered summary of the transcript instead of the transcript itself
  • (for fun) extending the integration to create visuals or animations based on the transcript content

Built With

Share this project:

Updates