Interruption handling controls what happens when a caller speaks while the AI assistant is in the middle of delivering a response. Getting this right is important for making conversations feel natural — a good AI assistant should respond to interruptions the way a skilled human would: gracefully stop, listen, and adapt.
There are two key settings to understand: interruption sensitivity (how easily the assistant stops speaking when it detects caller speech) and interruption behavior (what the assistant does after being interrupted).
Interruption Sensitivity
Sensitivity controls how quickly the assistant stops talking when it detects audio from the caller. Higher sensitivity means the assistant will stop at the slightest sound — even background noise or brief acknowledgments like "uh-huh." Lower sensitivity means the assistant will continue speaking unless the caller says something substantial.
| Setting | Behavior | Best For |
|---|---|---|
| High | Stops at any detected audio | Conversational, back-and-forth dialogues |
| Medium (Default) | Stops when caller speaks a full word or phrase | Most use cases — balanced responsiveness |
| Low | Continues unless caller speaks for 1+ seconds | Delivering important information that should not be cut off |
Handling Background Noise
A common issue with high sensitivity settings is that background noise — a TV, traffic, or other people talking — can trigger interruptions even when the caller is not actually trying to speak. If you notice your assistant stopping frequently due to background noise, try reducing the sensitivity setting or enabling noise suppression in your assistant's audio settings.
Related Guides