onepagers is a set of focused PDF extraction tools. Each one is built for a specific document family and keeps extracted facts connected to their source.
The current tools cover Closing Disclosure and Loan Estimate PDFs, legal discovery productions, and K-1 or 1099 tax forms.
It finds loan amount, interest rate, monthly principal and interest, closing costs, cash to close, term, and lender name in supported mortgage disclosure PDFs.
It builds a first-pass index of dates, parties, dollar amounts, exhibits, and Bates references, with the source file and page beside each result.
It recognizes supported K-1 and 1099 layouts, maps their labeled box values, and shows the source box, page, confidence, and text used for each field.
The tools differ. Discovery can use OCR on pages without usable text; the mortgage and tax extractors require a selectable text layer, so run OCR before uploading an image-only PDF to those tools.
Yes. Depending on the tool, results include the source file, page, form box, Bates reference, nearby text, or a highlighted source region.
They describe how strongly the extractor matched a value to its expected label, box, or source text. Use them to prioritize review, not as a guarantee.
The K-1 Extractor lets you edit values in its review table and export the checked rows. The other tools provide their own result exports without an in-page correction table.
The upload controls accept PDF files. They do not accept Word documents, images, spreadsheets, or other file types directly.
Free processing is capped at 8 MB and 15 pages across the tools. Discovery accepts batches up to 50 PDFs and one PDF up to 80 pages; K-1 allows up to 25 pages, while the mortgage extractor uses its free 15-page boundary.
Discovery accepts up to 50 PDFs in one batch. Closing and K-1 handle one PDF per review.
Files over a free-processing boundary are accepted and queued for heavy processing where that tool supports it. Discovery rejects more than 50 PDFs or a PDF over 80 pages, and K-1 rejects PDFs over 25 pages.
Uploaded PDF bytes are deleted after processing. The active review keeps extracted results and source metadata for the current session; the tools do not provide a document history.
No. The current product copy states that uploaded documents are not used to train models and processed file contents are not made available to staff.
The public experience is the browser upload tool. A public, documented API with developer signup and API keys is not advertised in this version.
Accuracy depends on the document layout, text layer, scan quality, and how closely the PDF matches a supported form. Review important values against the source document.
The browser tools are available to try under their free-processing limits. No paid-processing price or payment flow is published yet.
Yes. Depending on the tool, you can download CSV or JSON results, or a checked CSV from the K-1 review screen.
They are for people who repeatedly review mortgage disclosures, legal discovery productions, or tax forms and need structured facts without losing the source evidence.