Good question @AdeptSchneider22 .

@avt_nyanja @Zindi

very good question

We are looking for a combination of fine tuning and RAG. RAG is for the information retrieval bit. You are leveraging the broad knowledge of your LLM of choice by finetuning to understand and respond to disease survellance questions. With RAG, you should be able to update the knowledge base with other information pertaining to disease surveillance, e.g., changes to some guidelines and the model should be responsive to that.

I hope my answer is adequate. But in a nutshell you are the solution provider so you can choose and balance between RAG and finetuning as needed.