RAG knowledge store with inferencing, predictions, and question answering.

Using the attached files, upload pdfs to the Azure AI Search, create a Search Index with recognizable fields for extracting answers for questions surfaced to the client via queries to an LLM endpoint configured with RAG custom implementation or an agent file search tool.

RAG requirements include:

1. Foundry project
2. Azure AI Search
3. Upload pdf sources
4. Index, Vectorization, and Embeddings
5. LLM endpoint
6. Endpoint and index integration
7. Prompt edits for injecting updates from RAG knowledge base
8. Test queries with prompts, system message, and supporting content
9. Test chain of thought and few-shot learning
10. Source validation