We will optimise an LLM using 'Retrieval Augmented Generation' or RAG,
where we provide the model with additional knowledge with which it can
generate more grounded, relevant and accurate responses to our queries.