Based on that OpenAI-Pinecone webinar from last year, this is the screenshot I took of their user interface. It served as the model I would aspire to for the next several months:
And, below is the user interface I finally developed:
As you can see, citations are listed. There are also “library” selection options (groups and tags) as well as options to fine-tune some chat-completion elements (concepts, standalone question, hybrid search, etc…).
It was a long road, but I finally got there.