WorkDeliveredLive CaptureTechnical DesignMilestonesFull Pipeline

Core — Post-Meeting

Slice 4a Core.

Slice 4: Post-Meeting Reprocessing + Speaker ID

Owner: Core engineer Domain: Core Complexity: M Status: ✅ Done

The automatic pipeline that runs after recording stops.

Tasks

publish TranscriptionJobMessage and MeetingTerminatedMessage on session stop consume MeetingSessionTerminated — drain AssemblyAI buffer; send final segments via REST
ML: skip task extraction when existing live tasks present
Core: POST /transcriptions/{id}/remap-speaker and POST /transcriptions/{id}/identify-speakers
ML: speaker centroid computation + identify + remap pipeline

Test cases:

Test	Location	Assertion
Post-meeting pipeline triggers automatically	`test_system`	Recording stop → transcript segments replaced with higher-accuracy version
Tasks from live session preserved	`test_system`	After post-meeting processing, user-created and system tasks still present
Speaker labels resolved to person_id	`test_core`	Cosine similarity match → segment `person_id` updated
Talking points promoted to `is_final: true`	`test_system`	After post-meeting processing, talking points have `is_final: true`
Headline generated	`test_system`	Meeting has non-null `headline` after post-meeting processing

Full Pipeline Operational

Milestone 2 integration overview.

ML — Post-Meeting

Slice 4b ML.

On this page

Slice 4: Post-Meeting Reprocessing + Speaker ID Tasks