RAG System Setup with Weaviate and Prefect

Design and implement the RAG pipeline. Set up a **Weaviate** instance (e.g., via Docker). Use **Prefect** to create a workflow that: 1) takes a collection of sample policy documents, 2) chunks them into smaller, semantically meaningful passages, 3) generates embeddings for these chunks (using an appropriate embedding model, e.g., OpenAI's `text-embedding-3-small`), and 4) indexes them into Weaviate. Provide the Prefect flow definition and Python code for Weaviate interaction.