Extraction, shown on real documents
Each page carries a genuine sample, the schema used to process it, the output ADE returned, and the exact region of the page every value came from.
Consolidated 1099
One consolidated 1099 carries five IRS forms. A brokerage bundles 1099-B, 1099-INT, 1099-DIV, 1099-OID and 1099-MISC into a single statement, adds pages of supplemental detail the IRS never asked for, and sends tens or hundreds of pages to anyone with an active account. Fidelity, Schwab, Vanguard, Edward Jones and Robinhood each lay theirs out differently. Anyone can read the summary boxes. Underneath them sits the transaction detail: every sale, when it was acquired, what it sold for, what it cost, any wash sale adjustment, and whether the holding is short or long term, covered or noncovered. That is what a return is built from.
66-field schema · grounded sample
Investor Presentation
Investor presentations put a company's numbers in their best light, and almost none of those numbers sit in a table. Quarterly earnings decks, pitch decks and analyst day presentations carry their substance in bar charts, pie charts, footnotes and dense mixed layouts. They run landscape, and every issuer designs their own. A text layer gives you the axis labels in whatever order they were drawn. Reading the slide gives you the series, the categories, the footnoted adjustments, and how they relate. On a deck, that difference is the whole job.
23-field schema · grounded sample
Invoice
Most companies process more invoices than any other document, and no two suppliers format them alike. One accounts payable inbox carries purchase invoices, credit notes, proforma invoices, utility bills and freight invoices. They arrive as native PDFs, scans, phone photographs and email attachments, in layouts that change without warning. Reading a total is easy. The work is everything around it: a supplier you have never seen before, a line item table that runs across three pages, and an auditor who wants to know where a number came from six months later.
35-field schema · grounded sample
