A Generative AI Engineer is tasked with building an LLM-based question-answering system that needs to handle newly published documents on a regular basis. The engineer wants to minimize both development effort and operational costs. Which combination of components and configuration will best meet these requirements?