> ## Documentation Index > Fetch the complete documentation index at: https://docs.parley.so/llms.txt > Use this file to discover all available pages before exploring further. # Tables > Ask the same set of questions across hundreds of documents at once and get structured answers back. A Table asks one set of questions across a whole set of documents and returns the answers in a grid. Each row is a document or a matter, each column is a question, and every cell is the Agent's answer to that question about that row. It's the tool for the work that's too big to read one file at a time and too specific to search for. ## What firms use them for * **Evidence evaluation** — run the criteria for a case type across every exhibit and see which ones actually carry weight * **Intake review** — pull dates, employers, and status from a stack of client documents at once * **Portfolio checks** — ask the same compliance question across every open matter * **Wage and HR data** — extract structured figures from client spreadsheets and filings ## Building one Add the documents or matters as rows, then define your columns as questions. Write columns the way you'd ask a person: specific, one thing at a time, with the form of the answer implied. A column like *"Date of the employment offer letter, or blank if there isn't one"* returns something you can sort. *"Employment info"* does not. ## Working with results * Hover over any cell and click the pencil to edit the value directly in Parley before exporting. * Sort and filter to find the rows that need attention. * Export to CSV or XLSX when the table is the deliverable, or feed the results into a draft. ## Credits Tables draw credits per run, and cost scales with the number of rows and the complexity of the columns. A wide table across hundreds of documents is a real piece of work — worth a look at the [Usage](https://app.parley.so/settings/usage) page after your first few. Adding the documents themselves is always free. See [How credits work](/plans-and-billing/how-credits-work). Build the table on a handful of rows first and check the answers. A column that's ambiguous on five documents will be ambiguous on five hundred, and it's cheaper to find out early.