onyx is good for this, it is standard doc ingestion -> chunk -> embedding -> index -> query -> rerank -> answer.

there are a few other local apps with simple knowledge base type things you can use with pdfs. cherry studio is nice, no reranking though.