Hello callin.io Community,
I'm currently integrating VAPI.ai with a callin.io workflow to manage booking requests via a webhook.
The workflow is structured as follows:
- Webhook (receives incoming VAPI tool-calls)
- Check Google Calendar Availability (GetAvailability node)
- ** Ai Agent
- IF Node (determines if the requested time is available)
- Respond to Webhook (sends the result back to VAPI)
What I've Already Verified:
- toolCallId is correctly captured from the Webhook input (dynamic, not hardcoded).
- The response format precisely matches the VAPI documentation's requirements.
- Client messages (VAPI) include function-call and function-call-result.
- Server messages (VAPI) include tool-calls.
My Questions:
-
Why does VAPI not immediately speak the text from my webhook response?
-
Is there a specific element I need to include in the webhook response for VAPI to interpret it as text-to-speech?
-
Do I need to alter the response structure (e.g., use plain text instead of a complete object)?
Additional Details: -
I've experimented with various response types (simple text, structured JSON).
-
When I manually initiate a follow-up request, VAPI occasionally speaks, but not directly after the Respond to Webhook step.
-
It seems VAPI might be expecting a different structure or a specific flag within the webhook response.
Any insights or examples on how to resolve this would be greatly appreciated!
Thanks in advance!
Vapi: