https://vapi.ai/ logo
Join Discord
Powered by
missing end-of-call server POST
# support
a
Our server response to 'assistant-request' contains this:
Copy code
"serverMessages" => [
                        "end-of-call-report",
                        "status-update",
                        "hang",
                        "function-call",
                ],
but no 'end-of-call-report' gets posted. What's missing please? CallID e.g. 2241a7a3-3a98-44ec-b6e9-5a0387507976
v
Hi @AiBizBox, I'm an AI assistant for Vapi. I'll analyze available sources and get back to you shortly.
----------------------------- **While I generate an answer, here are some sources that might be helpful**: * [Create Assistant]() * [Get Assistant]() * [List Assistants]() * [Update Assistant]() -----------------------------
### Answer It sounds like there might be an issue either with the configuration on your server or on Vapi's end that is preventing the 'end-of-call-report' from being posted. First, ensure that your server is correctly set up to handle incoming POST requests at the URL you've configured in your Vapi assistant settings. This includes being able to handle the specific 'end-of-call-report' data format that Vapi sends. You can find more details about the data format for an 'end-of-call-report' in the Server Events documentation [(1)]() . If your server setup appears correct, and other server messages (like 'status-update' or 'hang') are being successfully received, this might indicate a more specific issue with the 'end-of-call-report' event. This could be due to a misconfiguration in how the assistant is set up to send events to your server, or it could be an issue on Vapi's end. To troubleshoot further, you can: 1. Check the assistant settings in your Vapi dashboard to ensure that 'end-of-call-report' is indeed enabled. 2. Review the server logs for any errors or warnings that could indicate why the 'end-of-call-report' is not being sent or handled correctly. 3. If possible, test with a simplified configuration to rule out any complex interactions that might be affecting the 'end-of-call-report'. If the issue persists after these checks, you may need to reach out to Vapi's support with specific details (like the call ID you provided) for further investigation.
Got it. Tagging @User for additional help.
a
j
have you set up the "serverUrl" parameter inside your assistant?
a
no (as you can see from the screenshot above), the phone number is using the our DEFAULT ServerURL. And that's working perfectly fine for 'assistant-request' POSTs. However, no 'end-of-call-report' POSTs are coming thru. @User any idea maybe?
j
You need to configure the serverUrl within the assistant so it knows where to send the end-of-call report to. It doesn't automatically use your default serverUrl to send the end-of-call report. It only sends the assistant request to your default serverUrl
a
I've got this configured, so how to additionally configure the serverUrl please? And as a side note: I find that counter intuitive - why wouldn't it send the end-of-call report to the default server url as well? https://cdn.discordapp.com/attachments/1242123278789906554/1242197649973837834/image.png?ex=664cf636&is=664ba4b6&hm=03b4a94c97ae687d62787358030301a756e2d5df1f58fabb69854f35eb08fc7e&
the SAME url where it sent the 'assistant-request' to (and received the config above from)
ok I've added ``"serverUrl" => "https://ourServer.com/enpoint.php"`` but that doesn't help. Instead we're getting a strange behavior where the AI is no longer waiting for any answers from the caller but instead uses answers given on a previous test call to answer itself and proceed in the script without any interruption. Totally weird. id b71a12a6-5997-4496-81bb-8148b0a23d90 @Sahil https://cdn.discordapp.com/attachments/1242123278789906554/1242221156279255070/image.png?ex=664d0c1b&is=664bba9b&hm=d6bbc8a57758b9761b396903f6e7432111b1990b0f17eece0d52617b6a0952ca&
that's with mixtral-8x7b-32768 on groq
s
can you sue gpt-4o probably and check?
a
not sure what you mean by 'sue' - could you please explain in a bit more of detail.
s
typo.... I meant can you try gpt-4o and lmk if you are still facing the issue?
a
gpt-4o partly solves the issue: it still reads out short example answers like e.g. [yes] or [ok] from the prompt, but at least it now waits for the caller to say that (too). On longer example answers it's working fine with gpt-4o. Please listen in yourself on id d2abb647-fe30-4e83-b4e5-0c4cc9ee4502. Also, still not getting the end-of-call-report POST.
still not getting ANY end-of-call POSTS. @Sahil what do you recommend testing next? Explicitly setting ``serverUrl`` (see screenshot) didn't have any effect unfortunately. https://cdn.discordapp.com/attachments/1242123278789906554/1242463054265061507/image.png?ex=664ded63&is=664c9be3&hm=8409e8887148c41533a80ffaa8f8f125f637fc727ace85768bf6f3d67c944f52&
s
Can you send me the payload once to DM i will check it myself.
a
done 🙂
s
I am getting the end-call-report using your payload.
PreviousNext
Powered by