Showcase agents / Voice Appointment
Healthcare schedulers
Voice Appointment
What problem? Voice pipelines fail opaquely — teams cannot tell whether latency is STT, LLM, or TTS without stage-tagged telemetry.
Voice reference: STT → dialog → book/reschedule → TTS. Stage latency, failover paths, and per-stage SLO attribution.
What Inferix observes
Every run dual-writes to LensAI and TraceForge. Switch to this tenant in the console to see live KPIs.
- LensAI — P99 latency by stage (stt | llm | tts)
- TraceForge — stage waterfall across the voice pipeline
- DriftWatch — dialog quality vs golden transcripts
- RouteIQ — failover routing on STT latency injection
Tools & models
Real SQLite backends, not mocked HTTP stubs.
- stt.transcribe
- calendar.availability
- calendar.book
- calendar.reschedule
- crm.update_appointment
- tts.synthesize
Model routing
- STT service
- RouteIQ-routed LLM
- TTS service
- failover paths
Try it
Start the stack, then run the golden scenario. Traffic appears in the console within seconds.
1. Start stack + agents
cd inferix && ./run-inferix.sh --full --agents
2. Golden scenario
curl -s -X POST http://localhost:9105/scenarios/book_new | jq
Expect HTTP 200 with JSON status and tool steps. Traces appear under this tenant in ~10s.
3. All agents at once
cd inferix && ./scripts/generate-agent-traffic.sh
Expected: JSON with status and tool steps. Then open the tenant console and filter traces by tenant-voice-health. In-browser runner stays deferred — curl is the verified path.
Console: /console/overview?tenant_id=tenant-voice-health · API traces: GET /v1/traces?tenant_id=tenant-voice-health
See all seven agents in the console
Multi-tenant super-dashboard. Filter by tenant, drill into traces and costs.