### Answer Yes, Vapi can indeed act on responses from custom functions. When you define a custom function, you can configure it such that Vapi waits for a response from the function and then continues the conversation based on that response. This is facilitated through the use of a

serverUrl

where the function call information is sent, and the response from your server can influence the subsequent actions or prompts of the assistant [(1)]() , [(2)]() . Here's an example of how a function might be set up to send an email and act based on the response:

json
{
  "functions": [
    {
      "name": "sendEmail",
      "description": "Used to send an email to a client.",
      "parameters": {
        "type": "object",
        "properties": {
          "emailAddress": { "type": "string" }
        }
      }
    }
  ]
}

When this function is triggered, Vapi sends the function call to the specified

serverUrl

. The assistant can then process the server’s response to determine the next steps in the conversation. This setup is particularly useful for creating dynamic and responsive voice applications where real-time decision making is crucial based on external data or operations.

Got it. Tagging @User for additional help.