DocExtract

Client-side document extraction. Upload an image, OCR it with Tesseract.js, and extract structured fields using a predefined schema.

How it works: OCR Only mode runs real Tesseract.js in your browser on uploaded images. Vision mode uses simulated AI results. The "Use Sample Invoice" button demonstrates extraction against hardcoded sample text. This is a simplified version of the production IDP System's extraction pipeline.

OCR MODE

EXTRACTION SCHEMA

Vendor NametextInvoice NumbertextInvoice DatedateTotal AmountnumberCurrencytext

Drop an image here, or click to browse

Supports PNG, JPG, TIFF, BMP