Embedded Document Processing

Accept complex documents from your customers and return structured data through one configurable service.

The Reference Flow

A document enters through the API or a channel you control, is pre-processed, classified, split, and extracted, then returned as structured data with per-field confidence scores. A webhook fires on completion. See the underlying data extraction engine for how context-based reading works.

What It Handles

• Channels: API, email, SFTP, shared drives, portals, and file sync.

• Formats: PDF, images, spreadsheets, forms, scans, photographs, dot-matrix prints, and mixed packs.

• 300+ languages with native Asian script support, plus handwriting and rubber stamps.

Configuration and the Review Loop

Model Builder defines fields with drag-and-drop, editable or locked at model level, with data types per field and per line-item header. Queue rules handle validation, set-values, and routing. Fields below the configurable confidence threshold, default 0.9, route to review. You surface Staple's review step or build your own on the API.

Proven In Production

A European KYC technology provider onboarded new document types across multiple languages with no code. A global FMCG brand processed supplier documents in four APAC languages, including dot-matrix prints a legacy OCR tool could not read, at 99.6% accuracy.

Scope

Packaging is usage-based and scoped per deployment. You decide whether models are managed centrally or configured per tenant.

Talk to the Platform Team

Discuss your document types and tenant model, or view the API documentation.