docs(live-stt): semantic end-of-turn feature page#128
Conversation
Documents the end_of_turn config for live transcription and, most of all, explains endpointing (time/silence-based) vs semantic end-of-turn (meaning-based) so users know which to reach for. - new chapters/live-stt/features/end-of-turn.mdx: concept, the endpointing-vs- end-of-turn distinction, parameter reference (threshold, emission_threshold, max_wait_secs, probe_base_ms) in user terms, and when to use which - endpointing.mdx: a short "Endpointing vs end-of-turn" section cross-linking the two - docs.json: register the new page in the Live Transcription nav Framed purely on user-visible behaviour (latency, turn-taking, premature cuts); no engine internals.
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (3)
📝 WalkthroughWalkthroughAdds a Live Transcription chapter for semantic end-of-turn detection, documents its configuration and relationship with endpointing, and includes the page in the documentation navigation. ChangesSemantic end-of-turn documentation
Estimated code review effort: 1 (Trivial) | ~3 minutes Possibly related PRs
Suggested reviewers: 🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
|
Preview deployment for your docs. Learn more about Mintlify Previews.
💡 Tip: Enable Workflows to automatically generate PRs for you. |
|
Preview deployment for your docs. Learn more about Mintlify Previews.
💡 Tip: Enable Workflows to automatically generate PRs for you. |
1 participant
Adds a Live Transcription docs page for semantic end-of-turn, and — the main goal — explains endpointing vs end-of-turn so users know which to reach for.
What
chapters/live-stt/features/end-of-turn.mdxthreshold,emission_threshold,max_wait_secs,probe_base_ms— in user termsendpointing.mdx: short "Endpointing vs end-of-turn" section cross-linking the two pagesdocs.json: register the new page in the Live Transcription navFraming
Purely user-visible behaviour — latency, turn-taking, premature cuts on hesitation. No engine internals, no benchmark numbers (those live with the feature work, not the docs).
Pairs with the
end_of_turnstreaming config in the API.Note
Low Risk
Documentation-only changes with no runtime, API, or security impact.
Overview
Adds Live Transcription documentation for semantic end-of-turn, aimed at helping users choose between silence-based endpointing and meaning-based turn finalization.
The new
end-of-turnpage explains the feature for voice agents and IVR, contrasts it with endpointing (including a comparison table), documents enabling via theend_of_turnlive config object, and describes user-facing knobs (threshold,emission_threshold/speech_turn_end,max_wait_secs,probe_base_ms) plus guidance on when to use each approach.endpointing.mdxgains a short Endpointing vs end-of-turn section that links to the new page, anddocs.jsonregisters the page in the Live Transcription nav after endpointing.Reviewed by Cursor Bugbot for commit b56b2bf. Bugbot is set up for automated code reviews on this repo. Configure here.
Summary by CodeRabbit
New Features
Documentation