Play 39
AI Meeting Assistant
Meeting companion — transcription, speaker diarization, action items, decision tracking.
Intelligent meeting companion providing real-time transcription, speaker diarization, automatic action item extraction, decision tracking, follow-up scheduling, and meeting summary generation. Azure AI Speech handles transcription and speaker identification, Azure OpenAI extracts structured insights, Microsoft Graph integrates with calendar and email for automatic follow-up scheduling, and Container Apps hosts the processing pipeline.
Architecture Pattern
Meeting AI: transcription → diarization → extraction → scheduling, Graph integration
Azure Services
DevKit (.github Agentic OS)
- agent.md — root orchestrator with builder→reviewer→tuner handoffs
- 3 agents — Meeting Builder (gpt-4o), Reviewer (gpt-4o-mini), Tuner (gpt-4o-mini)
- 3 skills — deploy (167 lines), evaluate (181 lines), tune (233 lines)
- 4 prompts — /deploy, /test, /review, /evaluate with agent routing
- .vscode/mcp.json — FrootAI MCP with Speech + OpenAI key inputs + envFile
TuneKit (AI Config)
- config/openai.json — meeting analysis prompts
- config/meeting.json — action item rules, follow-up scheduling
- config/guardrails.json — PII in transcripts, consent tracking
- evaluation/eval.py — Action extraction >85%, Summary quality >80%
Tuning Parameters
Machine evidence
FrootAI evidence lifecycle
This is an internal evidence maturity label, not third-party certification, accreditation, legal compliance, or a production guarantee. Missing or expired evidence demotes automatically; catalog claims cannot promote a play.
This play currently has design evidence only. A runnable scenario, endpoint evaluation, and build receipts are the next contiguous gates.
Repo Intelligence
v1A no-clone, revision-pinned map for agents and humans. Observed evidence is separated from inferred flow so the output stays useful without pretending to be a full call graph.