Skip to content

Local audio/video transcription #21

Description

@ai-mindset

Add support for local audio/video files (mp3, mp4, wav, m4a)

Currently Synthesis only ingests YouTube URLs via yt-dlp subtitle extraction.
For webinars and recorded meetings stored locally, we need to transcribe
audio/video files directly.

  • Implement a LocalMediaFetcher using the existing FetcherBehaviour
  • Integrate with local Whisper via Ollama for transcription
  • Support: mix wiki.add /path/to/meeting.mp4

This keeps all data on-prem and removes YouTube dependency for internal content.

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or request

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions