The AI Receptionist is currently dropping caller context due to a hardcoded platform-side fallback prompt. The system autonomously injects the phrase, "Please let me know how I can assist you. I'm here to help." during brief processing pauses (e.g., while the model is fetching a Knowledge Base result) or occasionally immediately after the initial greeting before the caller has a chance to respond.

Currently, there is no way to disable this generic fallback prompt or extend the silence-timeout threshold from the Admin Portal.

Business Impact
When this hardcoded prompt fires prematurely, callers assume the AI has reset or lost their context. This creates a poor user experience, resulting in callers hanging up or having to repeat their requests (e.g., having to say "Speak to a representative" multiple times before a transfer initiates).

Requested Enhancements (In order of customer preference)

Configurable Silence-Timeout: Add an Admin Portal setting to adjust the silence-timeout threshold per AI Receptionist. The current default appears to be ~3-5 seconds; the customer requests the ability to extend this to 10-15 seconds.

Admin Portal Override: Allow administrators to customize, override, or entirely disable the generic "No Input" fallback prompt.

Mid-Generation Suppression: Suppress the fallback prompt entirely while the AI model is actively in a processing turn or mid-response generation.

Additional Notes
Providing partial relief by implementing Option 3 alone would meaningfully improve the customer experience and resolve the most disruptive element of this issue at their current daily call volume.