RAG (Geração Aumentada por Recuperação)

An AI architecture that combines LLMs with external knowledge retrieval. Instead of relying solely on training data, RAG systems retrieve relevant documents from a knowledge base (using embeddings and vector search), then provide them as context to the LLM. RAG reduces hallucinations and enables up-to-date responses. Useful for blockchain documentation bots and developer assistants.