How do you guys handle your personal RAG setup

I am getting into developing a RAG setup, for getting information out of existing documents, new document ingestion, web searches, and good visuals.

I am planning to use it for, alongside the regular "chat to my data", ingesting personal docs, invoices, creating tables views and recurrent jobs to handle updating those views.
I also want to have the least hallucinations possible, so i think i will need a real ocr services instead of just vision LLMs

i tried anything LLM previously, but it was super clunky and the UX wasn't as easy as i wanted to.

Is there any known solutions, or stacks that you have running or can vouch for ?