WordloopLive Recording Operational
Milestone 1 integration overview.
Milestone 1: Live Recording Operational
Integration Lead: App engineer (closest to the user) Combines: Slice 1 (Core) + Slice 2 (ML) + Slice 3 (App)
The user can start a recording, watch live transcription and AI insights stream in, add manual tasks, and stop the recording. The core value proposition of the bet is functional.
End-to-end tests:
| Test | Assertion |
|---|---|
| Full live recording flow | User starts recording → sees live transcript → sees talking points → adds task → stops recording → meeting saved |
| Multi-speaker recording | Recording with 2+ speakers shows distinct speaker labels on segments |
| Graceful degradation | With ML unavailable: recording continues, audio captured, degraded banner shown |
| Duration limit enforcement | After configurable limit, recording auto-stops and post-processing triggers |
| Concurrent session prevention | Starting a second recording while one is active → error shown |