Hi, Im using a vapi inbound bot, that is connected to make.com. When I call the bot now, it functions properly, but for some reason it sends 3 payloads during one call to make.com instead of just one. First I contacted make support, but they refered me to you. Im going to provide you the responce I got from them. : ´I have checked this deeply and I can see that Vapi sent 3 different payloads for the same call, here are some of the differences: unnamed.png So I believe VAPI would be in a better position to provide more clarity on what went wrong there.´ Thank you for your early responce. Adam Icsó

@Adm420419 Hello Adam, Thank you for reaching out with your concern. It seems unusual for the VAPI inbound bot to send three payloads for a single call when it's expected to send only one. To assist you better, I would need to check the configuration and logs on our end to understand why multiple payloads are being sent. Could you please provide me with the following details? 1. The time frame during which the issue occurred. 2. Any unique identifiers for the call(s) in question (such as call IDs). 3. The expected payload structure. With this information, I can investigate the matter more thoroughly and help pinpoint the cause of the issue. Once we have identified the problem, we can work towards a resolution. If you have any additional information or context that might help, please feel free to share that as well. Looking forward to your response so we can resolve this issue promptly.

My make scenerio has a router with a yes and no branch. When data passes trough yes branch, everything goes well. When data passes trough no branch, the bot (on the call) gets stuck, but when reviewing make logs, vapi passes 2 payloads, which in make go right and the scenario even gets finished. Example (phone appointment setter): Call begins, bot suggest first date, when declining, it gets stuck and again suggeste the first day but it also starts mixing words. In make though, it finds the next date, scenario passes, but a few seconds after another payload comes and suggest and passes the next date. Example: 03.06 - 14:00 gets declined, 04.06 - 11:00 gets suggested - passes in make without my confirmation in the phone and a few seconds after 04.06 - 11:15 gets sugested and also passes, also without my confirmation, during the same call.

@Adm420419 It sounds like there might be an issue with the logic handling within the VAPI bot or a misconfiguration that's causing it to send multiple payloads. This could be due to the bot not properly handling the "no" branch, leading to unexpected behavior and multiple payloads being sent to Make.com. To address this issue, we should review the VAPI bot's configuration, specifically the logic that determines how to handle declined suggestions and how it interacts with the Make.com scenario. Here's what we can do: 1. **Review the VAPI bot logic**: Check the conditions and actions set for the "no" branch to ensure they are configured correctly and not causing the bot to loop or send additional payloads. 2. **Check for errors or warnings**: Look for any errors or warnings in the VAPI logs that might indicate why the bot is getting stuck or sending multiple payloads. 3. **Test the scenario**: Run a controlled test with the bot and monitor the behavior when the "no" branch is triggered to see if the issue can be replicated and to gather more information. 4. **Consult documentation**: Ensure that the bot's behavior aligns with the VAPI documentation and that there are no known issues with the current version you are using. 5. **Contact VAPI support**: If the issue persists and you're unable to identify the cause, reaching out to VAPI support with detailed information about the problem, including logs and configuration details, would be the next step. If you can provide me with the specific details of the bot's configuration and any relevant logs, I can assist you further in troubleshooting the issue. Otherwise, please consider the steps above to investigate and resolve the problem. If you need any assistance with these steps or have any other questions, please let me know. I'm here to help!

@Vapi I want to talk to a human about this issue

@Adm420419 Please ask support for more info on this.

@Vapi how?

@Adm420419 I apologize for any confusion. To speak with a human about this issue, you will need to contact VAPI's customer support directly. You can usually do this through the official website or customer support portal of the service provider. Look for options such as "Contact Us," "Support," or "Help Center." Here are some general steps you can follow: 1. Visit the official VAPI website. 2. Navigate to the Support or Contact Us section. 3. Choose the method of contact that suits you best – this could be a support ticket system, email, live chat, or phone support. 4. Provide all the necessary details about your issue, including what you've experienced, steps you've already taken, and any relevant identifiers or logs. If you have access to a customer portal or dashboard, you may also find options to contact support there. If you need further assistance in reaching out to support or have any other questions, please let me know, and I'll do my best to assist you.

@Adm420419 you might have enabled few of the server messages which your refering as payload, please share the call id.

055a9267-a6d5-4ffb-90da-3ace6ecb9472

hey can please share the timestamp.

Assistant speech started at 13:23:33:574 (beginning of the call) and at 13:24:14:390 it started again (after the responce was ´no´) the first make scenario started at 13:23:54 and the second one at 13:24:14

hey your saying about the tool calls? if no then ping me once your back.

Yes, the timestamps I provided are related to the tool calls. Specifically: Assistant speech started at 13:23:33:574 (beginning of the call). It started again at 13:24:14:390 (after the response was 'no'). These are the moments when the bot's assistant speech began and when it restarted after I declined the first date, which shouldn´t have happend. Please let me know if you need any additional information. Thank you!

@Adam do you have other call id as well, I couldn't make my mind something is wrong.

Sorry i didnt understan you. You want me to provide another call with timestamps?

I tried looking into transcription and prompt but couldn't reach to the conlusion so i require another call id.

The issue happens the same every time. its after when the first suggested date gets declined during a call. The issue is that during the call, after the caller declines the first suggested date ´Assistant speech started´ starts again, which as far as i know, it should only happen once during a call and that is at the beginning

@Adm420419 check DM.