Comment by srijansriv

This, specially the nlp part, strikes me as a primitive version of the RAG-LLM pair