How do you guys handle your personal RAG setup

A developer is seeking advice on setting up a personal RAG (Retrieval-Augmented Generation) system. The goal is to extract information from existing documents, handle new document ingestion, perform web searches, and generate good visuals. The user plans to use it for "chat to my data," ingesting personal documents, invoices, creating table views, and recurrent jobs for updates. To minimize hallucinations, a real OCR service is preferred over vision LLMs. The developer is looking for known solutions or stacks that others can recommend.