A look inside the LLM closed box: test, observe and evaluate your RAG assisted chatbot Similarity score = 0.60 More