ONE Voice Lab · Experimental · Simulated
Simulation Lab
Change bounded synthetic conditions, watch failure propagation, and compare evidence without pretending it is production telemetry.Discovery only · replay inactiveNo provider spendOne run is not a benchmark
Scenario registry
Stress the system deliberately
11 defined scenariosFlagship experiment - Implemented
Target Speaker vs. The World
A flagship deterministic replay for target speech, crosstalk, environmental noise, interruption, silence, and recovery.
ObserverDiscovery only
- Audio
- STT
- Conversation
- Agent
- Tool
- TTS
- Playback
- Outcome
Realtime view
Event timeline
No run is active
Opening this page made zero provider requests. Scenario and template inspection remain available; replay execution is inactive.
Owner-controlled usage
Live provider mode remains gated
This installation permits discovery and existing evidence inspection only. New synthetic runs and live provider execution are unavailable.
- Provider requests
- 0
- Audio submitted
- 0 sec
- Billing
- Usage captured. Billing value not estimated.