OpenAI Assistant Creating New Vector Storage

Goal: Invoke an assistant with a pre-loaded vector store and receive the response within my flow.

Problem: When I attempt to call the assistant with an already assigned vector store, it generates a new vector store and indicates that it's empty, thus preventing the completion of my search.

Steps:

1. Verified functionality in the OpenAI playground; it operates correctly there.
2. Examined the saved threads between the playground and callin.io scenarios. When invoking from the playground, it accesses the correct vector store and delivers the necessary results. However, when calling from callin.io, it creates a new vector store and fails to return the required results.

Could there be an issue with the assistant module that causes it to automatically create a new vector store when a query is made to the assistant?

image1011×1054 45.6 KB

image2122×940 21.8 KB