I have successfully setup a ChatGPT driven AI Assistant using MS Teams, MS Power Automate and OpenAI API Endpoints (i.e. Create Thread, Create Message, Run and Retrieve Thread once complete). I am now trying to switch the API endpoints from OpenAI to my Azure Deployment/Endpoints.
There are some small adjustments I needed to make to the HTTP calls and headers and these have been successful apart from the last one - Retrieve thread (GET).
Does anyone have any pointers on what I may be doing wrong?
The response is not providing the response from the OpenAI service to the prompt in the output (see below).
Also, I have requested an update to the quota, but have also tried other models/deployments and it’s still not working? Is it a quota issue? There’s no error suggesting this just fails at the step after the HTTP Request because the data is null.
GET
https://myresource/openai/threads/{my threadid}?api-version=2024-05-01-preview
Thanks Savin, that’s a good idea, I’ll adopt the variable approach using .env. However, even when I’m hardcoding the API endpoint in (using a threadid variable) to retrieve the completed run, the JSON I get back is very short and does not have the data:tag included with the thread messages embedded as it does in the OpenAI (non-Azure) endpoint.
All the other REST API endpoints return the right information, just not this one. I feel like I’m missing something basic?