Hi Fabio,

Thank you for the feedback. By that I meant that you can make use of the LLMs large context window...which implicitly lets the model perform a RAG type process internally.

Or RAG can be setup outside of the LLM and only a smaller chunk of data submitted to the LLM.