I have set up a database that retrieves user information during their calls (name, surname, etc.). The goal is to send this data to VAPI via the API, specifically through the Initiate-VAPI function, so that the UDAPT assistant can remember the user and continue the conversation seamlessly during the next call. However, I am facing a major issue: While the data is correctly sent to VAPI via the API, I do not receive any logs or confirmation that VAPI has successfully received this information. Additionally, during the next interaction with the assistant, it does not remember the user, which suggests that the information sent to VAPI is not being properly stored or processed. Questions: Is there a way to get logs or confirmation from VAPI when sending the data via the API? Are there any specific parameters or configurations that need to be adjusted so that VAPI can properly "remember" the user for future interactions? Are there any solutions to resolve the issue of the assistant not remembering the user, or is there an alternative approach to ensure that the conversation continues where it left off? I’m open to any suggestions or adjustments that can help ensure the integration works properly.

Sorry about the trouble. There aren't any call logs at all from the API in your dashboard? Can you please provide your Org ID for your account so we may investigate? For agents "remembering" users, you could try using call analysis to extract information from the previous calls:

Hello, For the past 3 days, I’ve been trying to figure out how to integrate some data from my database into the VAPI prompt system. I’ve tried several approaches, but nothing seems to work so far. My goal is to send this data upfront, as soon as the user initiates the call, so that my assistant immediately has access to the following information: Name Age Origins Important info Summary of the last 5 discussions Please note that I’m not using phone numbers — all calls are made directly from my web app, and users are identified only through their user ID. As for the database, it’s quite straightforward: I internally retrieve the transcripts, analyze them using an LLM, and get the data I need in JSON format. Everything is stored in my backend. However, I’m still encountering the following error when trying to send this data to VAPI: ❌ Error sending history to VAPI: {"message":"Cannot POST /assistant/47a9ad21-7c41-4efd-8e27-21b63925e27b/history","error":"Not Found","statusCode":404} Thanks in advance for your help — happy to share more technical details if needed!

I'm assuming that what you're trying to POST is the history JSON file, but how does your integration have visibility of the transcripts that you need to send? I understand that it's stored in the backend, but if its a database connection I want to know how its retrieving the stored history