Parse the Folder, Not Just the PDFs: The Relational Tables RAG Needs on a Case File

When processing a complex case file, like an insurance claim, simply loading all the documents into a large context window is ineffective for answering the most critical questions. A case file is a bundle of dissimilar documents, and a user's primary needs are often to identify missing files or compare specific data points across different reports. A more powerful approach involves parsing the entire folder by first defining a list of all documents that *should* exist for that specific case type. This method allows the system to intelligently report on missing documents and perform crucial cross-document consistency checks, such as verifying that a date on a claim form matches one in an adjuster's report. Ultimately, the goal is to determine the overall state and completeness of the case, rather than just retrieving a passage of text.