persona-garden-patch

• is_a::[[Form Type]]
• has_status::[[Seed Stage]]
• in_domain::[[Deep Context Architecture]]
• in_precinct::[[Household Precinct]]

Transcribed Stage

A transcribed meeting has a text transcript produced from the recording. The spoken content is now machine-readable and searchable, but the transcript is raw — speaker labels may be wrong, filler words clutter the text, and formatting is whatever the transcription tool produced.

Transcription is the step that unlocks meeting content from audio/video into text. The quality varies by tool: Whisper produces bare text with timestamps, Fathom adds speaker labels and AI summaries, Otter provides real-time transcription with varying accuracy. Regardless of source, the raw transcript needs editorial attention before it can serve as a reliable reference.

Characteristics

• Full or partial transcript exists as text
• Speaker labels present but may be incorrect or inconsistent (first name vs full name, “Speaker 1” placeholders)
• Filler words present (um, uh, you know, like, so)
• Formatting reflects transcription tool output, not editorial structure
• Content is searchable — a significant step up from captured stage
• May include AI-generated summary from the transcription tool (Fathom, Otter)

Transitions

Transcribed → Cleaned: When the transcript has been processed through cleanup — speaker names normalized, filler words calibrated (removed or reduced to natural rhythm), formatting standardized. The /transcript-cleanup skill handles this step.

Sources

Processing track design from [[Linear Processing Stages for Meetings]]↑. Track placement within the three-track status system from [[Status Lifecycle Tracks]]↑.

Relations

• relates_to::[[Captured Stage]] — the previous stage in the processing track
• relates_to::[[Cleaned Stage]] — the next stage in the processing track
• relates_to::[[Status Lifecycle Tracks]]↑ — this stage belongs to the processing track

This site is open source. Improve this page.