My goal is to be able to implement it using the gpt-3.5-turbo intelligence model. Regarding the basic source code of the integration of this model, I did not encounter any problem. I stumble at the integration level for customer support precisely: memorizing conversations, training the AI as an assistant for a specific company, overcoming the openai API token limit problem.
I have another concern about this. How could I overcome the maximum context size length which is 4097 tokens. how do some apps like https://poe.com/ not have such a context limit?
That is done using embeddings.
Have you considered looking at a platform like CustomGPT that does everything you mentioned?