talk-stage1-extract

To structure raw notes and transcripts into a talk summary, provide your source file and optional metadata like event, date, and audience. The system extracts a narrative arc, main themes, and measurable metrics, flagging any content gaps to produce a markdown summary.

Yes, you can extract presentation metrics and narrative arcs from long-form articles. The process detects whether the source is experience-based or conceptual, then adapts the extraction to output 3–7 main themes alongside measurable metrics with their original sources.

To prepare a conference talk from source material, you should provide a slug, event name, date, duration, audience, and type. This metadata ensures consistent talk records and writes a structured markdown summary file to your talks directory.

Yes, the talk extraction process works for both experience-based and conceptual content. It auto-detects the source type and adapts the narrative extraction accordingly, ensuring the generated summary accurately reflects the original material's structure.

The best way to audit existing presentation material before building slides is to run it through a talk summary extraction. This identifies clear narrative arcs, quantifiable metrics, and content gaps, providing actionable recommendations for downstream pipeline stages.