Hi community,
I'm running into an issue with reliably starting a conversation as soon as a user answers a call. I have two main problems:
1. Silent User: If a user answers the phone and says nothing, my assistant (which is set to wait for the user) also stays silent, and the call eventually ends without any interaction.
2. Missed "Hello": Often, if a user says a short greeting like "Hello", the transcriber doesn't seem to catch it. As a result, the assistant never receives the first message and doesn't respond.
I've tested all three First Message Mode options:
* Assistant waits for the user: This is my preferred mode, but it causes the two issues described above.
* Assistant speaks first & Assistant speaks first with model generated message: These modes are not suitable because the assistant starts speaking too early, while the phone is still ringing.
My goal is to have the assistant start speaking only after the user has actually answered the call.
My question is:
How can I configure Vapi to reliably start the assistant's speech as soon as the call is answered by a human, solving both the "silent user" and the "missed short greeting" problems? Is there an event or a setting I'm missing?
Thanks for any advice!