A Generative AI Engineer must deploy a basic RAG application for a customer support system. The application requires embedding models for vector search, retrievers for document lookup, and an LLM for response generation. What are the steps to deploy this application?