How does one set up a live audio session with interrupts?

[consilience-dev]

Hi everyone,
I’m working on setting up a conversational voice application using ADK Web. I’ve run into an issue with implementing interrupts. Specifically, the model continues speaking even if the user starts talking again during the response.

From what I can tell, ADK Web doesn’t support this behavior natively. Are there any recommended approaches, workarounds, or example implementations for handling voice interrupts? Any guidance or documentation references would be greatly appreciated.

Thanks in advance!