Your question is Fine-Tuning vs RAG vs Prompting. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You are choosing how to adapt a large language model for a new application. You want a practical framework for deciding whether to rely on prompting, add retrieval with a knowledge base, or train a customized model.
Compare fine-tuning, RAG, and prompting. When use each?