Roadmap

Planned work items for the platform, organised by phase.

Phase 1 — Complete the Meeting Flow

The core capture → understand → extract → interact loop.

• Real-time transcript delivery to frontend (SSE / WebSocket)
• Live speaker diarization + identification (match to Person records)
• Live meeting summarisation (progressive LLM summaries)
• Auto-extracted action items from transcript → Task entities
• Sentiment / engagement scoring per transcript segment
• Live meeting UI (scrolling transcript, speaker labels, waveform, live insights)
• Meeting timeline / synced audio playback with transcript highlighting
• Command Palette enhancements for meeting navigation

Phase 2 — Platform Foundations

Infrastructure that everything else depends on.

• Service-to-service authentication
• User authentication (Auth0)
• Multi-provider transcription (Deepgram, Whisper as Gateway swaps)

Phase 3 — Beyond Meetings

See Future Vision for the longer-term product direction.

---

Future Vision

Placeholder — product direction beyond the core meeting flow. To be defined.