OCR for Documents

Convert W-2s and 1099s into machine readable text to improve automation and save your team’s time and energy.

Overview

Optical character recognition (OCR) converts uploaded or scanned W-2 and 1099 documents into machine-readable text, making them easier to review and unlocking auto-fill capabilities. This can accelerate your time to decide by bundling everything you need to move forward with Argyle, vs. needing to work with multiple vendors.

With webhooks, Argyle can notify you instantly whenever your user or their employer uploads a W-2 or 1099 on an ongoing basis, enabling you to make decisions based on the latest information available at all times.

This solution enables you to get data from W-2 and 1099 documents in two different scenarios:

  • When they are manually uploaded by your users during document upload.
  • When they are initially scanned by Argyle from a newly connected account, or when documents are added or updated through periodic scanning.
Two scenarios of data retrieval through OCR

To get the OCR for Documents solution enabled, reach out to your respective Customer Success Manager, Account Executive, or [email protected].

Testing OCR

Testing uploaded documents

Enable Document Upload using Link customizer and then use Link to upload either:

  • Argyle's sample PDFs that are provided in the Type-specific Reference page
  • Your own W-2 and 1099 documents in PDF format

Testing scanned documents

OCR for scanned documents is available by default. To test it, you only need to connect an account using Link.

Viewing OCR results

To see data retrieved through OCR processing, you can either list documents and list forms through the API, or check the Documents tab for a specific user within Console, where an OCR icon next to the uploaded or scanned documents will indicate the completion of the OCR process.

Supported document types

These document types are supported by OCR for Documents:

  • W-2
  • 1099

You can find two examples of ocr_data for the most common document types below.

Developer Tools

Use these reference pages to learn more about the API endpoints, webhooks, and form types used in the OCR for Documents solution:

FAQs

Paystubs are not currently supported.

Currently, only PDF files are supported.

OCR for Documents can achieve 97% accuracy.

15 to 20 seconds per document.