I am looking to develop a generative AI conversational application around file transfer analytics data. Users can come in and type question in natural language. The application understand users intent, generate SQL query, fetch data, do analysis on top of it if needed and respond back. Users should be able to ask more questions in a similar context, which would need llm to further answer questions about anomalies in data, analyze the cause/reasons of anomalies etc. In this process, it might need to talk to various tools like SQL generators, mathematical tools etc.
Please suggest the right framework and approach to handle this.
Maybe LangChain, have a look on the free course they have about it here!
@gent.spah Thanks. langchain seems to be the way. do you have any suggestions on how we can send data from the database for analysis? Since this is analytics data there are millions of records the LLM might need to analyse.
How are they doing it in the course, I cant remember about this actually, there might be another LangChain course there, have a look, or maybe a short course related to your question there are so many there!