Hello Bubblers, first of all, thanks very much for your help.
I am building an app that reviews transcripts of your calls and gives you feedback on how well did you. For example, it gives you a score and tells you which words you repeat throughout the transcript.
I am using the OpenAI API for this, and have done experiments before, and it works.
My question is, in your opinion what is the best approach to do this, keep in mind that I will be storing over +50 parameters of data that I will later use to generate the reports:
-I can prompt openai to return a JSON response and then I would utilize one of the many JSON PARSERS plugins to get individual parameters store in my DB
-Use the API approach where I would receive the response by openai and send that to bubble API to generate the JSON element and store that in the database.
-Use a pinecone database to store the information and then connect it to Openai to receive the parameters that I would store in the database.
My biggest concern is the amount of parameters I need to store and how to effectively manage them. I created workflows in the past using A JSON Parser plugin but I needed to manually input a parameter on each step and I don’t think this could be something manageable at this scale.
Anyway want to hear your ideas,
thanks in advance