meeting-autopilot

You can extract action items from meeting transcripts by processing VTT, SRT, or TXT files through a multi-pass LLM extraction pipeline. This converts discussions into structured outputs like decisions, tasks, and follow-up emails to drive execution.

Yes, VTT and SRT files can be processed to generate professional follow-up emails and ticket drafts. The pipeline parses transcripts to produce machine-friendly data and polished Markdown reports ready for copy, paste, or import.

Yes, you need access to an LLM API from Anthropic or OpenAI to convert meeting notes into structured reports. The pipeline leverages external LLM capabilities for multi-pass extraction of decisions and tasks.

The best way to track decisions and open questions across multiple meetings is by using a tool with local cross-meeting history storage. This allows you to store commitments and track open questions and parking lot items locally over time.

No, meeting transcript extraction requires jq and Python for robust data handling and local cross-meeting history storage. These dependencies are essential for processing VTT, SRT, or TXT inputs and managing the extracted data.

Transcript extraction may lack structured decisions and tasks if the input format is unsupported or the LLM API is not configured. The pipeline requires VTT, SRT, or TXT inputs and a valid Anthropic or OpenAI API connection for multi-pass extraction.