Play 84
Citizen Services Chatbot
Multi-language municipal services AI assistant with escalation.
Multi-language municipal services AI assistant with form filling guidance, appointment scheduling, permit status tracking, FAQ resolution, and human agent escalation for complex cases. OpenAI powers conversational understanding, AI Translator enables multi-language support, Communication Services handles channel delivery, AI Search indexes municipal knowledge bases, and Cosmos DB stores citizen interaction history.
Architecture Pattern
Citizen services pipeline: multi-language input - RAG knowledge retrieval - form guidance - appointment booking - escalation
Azure Services
DevKit (.github Agentic OS)
- agent.md — root orchestrator with builder→reviewer→tuner handoffs
- 3 agents — Citizen Builder (gpt-4o), Reviewer (gpt-4o-mini), Tuner (gpt-4o-mini)
- 3 skills — deploy (228 lines), evaluate (123 lines), tune (227 lines)
- 4 prompts — /deploy, /test, /review, /evaluate with agent routing
- .vscode/mcp.json — FrootAI MCP with OpenAI + Translator inputs + envFile
TuneKit (AI Config)
- config/openai.json - citizen interaction and form assistance prompts
- config/services.json - service catalog, language config, scheduling rules
- config/guardrails.json - escalation thresholds, PII handling
- evaluation/eval.py - Resolution rate >85%, Escalation accuracy >90%
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.