meeting-transcript

To clean ASR errors in meeting transcripts and generate analytical briefings, this Skill uses a two-stage pipeline: it first corrects raw transcripts with dialect detection and chunking, then produces per-company summaries using Gemini.

Yes, you can batch process meeting transcripts in .txt, .docx, and .pdf formats. The pipeline supports safe re-runs with skip-if-exists logic, retry mechanisms, and configurable retention thresholds for handling multiple files.

Yes, this meeting transcript cleaner works with Obsidian by generating Obsidian-ready frontmatter and per-company briefings that include backlinks and ticker extraction for cross-linking notes within your vault.

Yes, you need a Gemini API key configured via python-dotenv to automate meeting transcript summarization. The Skill depends on the google-genai library to perform the second-stage full-context briefing and analysis.

The best way to handle large meeting transcripts with speaker boundaries is using chunking at speaker boundaries during the ASR cleaning stage, which prepares accurate segments for the final Gemini briefing.