GPT API with LangChain and a Database
Budget: $250 – $750 USD
I've attached a file concerning what I'm looking for. Basically I'd like to have GPT (via GPTs API) have responses that can have templated elements that can be exchanged with information from a database. For example, a query may be made from the User to GPT as follows:
User: I haven’t gotten my order yet.
GPT: Your {order} information is {order_status}.
Where {order} and {order_status} is given by information queried from the database.
Now GPT has to determine whether the information requested is part of the website FAQ (after GPT reads the website) or a user specific query e.g., information specific to the user in a database. If it is determined by GPT that the query is asking about information, and it is in the FAQ, it then just gives a response based on its website reading. However, if it thought that the user wants particular information in their account, then a response is given with the information from the database included. More details in the attached file including a link to a partially worked out solution (I'm okay with you using GPT to help generate the full solution).
User: I haven’t gotten my order yet.
GPT: Your {order} information is {order_status}.
Where {order} and {order_status} is given by information queried from the database.
Now GPT has to determine whether the information requested is part of the website FAQ (after GPT reads the website) or a user specific query e.g., information specific to the user in a database. If it is determined by GPT that the query is asking about information, and it is in the FAQ, it then just gives a response based on its website reading. However, if it thought that the user wants particular information in their account, then a response is given with the information from the database included. More details in the attached file including a link to a partially worked out solution (I'm okay with you using GPT to help generate the full solution).