Overview
Welcome to Cardinal! We help AI teams and companies extract structured data from complex documents like PDFs, images, invoices, contracts, and medical records with industry-leading accuracy and security.
This guide provides an overview of our flexible toolkit of API endpoints, example use cases, and configuration best practices. Cardinal specializes in extracting not just text, but also bounding boxes, image metadata, barcodes, QR codes, signatures, and annotations from your documents.
If you have any questions unanswered, feel free to reach out to our team at team@trycardinal.ai
Core Features
These API endpoints are the building blocks for all your document processing needs. You can use them individually, or in conjunction to make powerful end-to-end OCR pipelines.
Output Formats
Cardinal supports three primary output formats, each optimized for different use cases:
Frequently Asked Questions
Cardinal goes beyond basic OCR to extract specialized document elements:
Machine-Readable Codes
- • Barcodes (Code 128, Code 39, UPC, etc.)
- • QR codes with decoded content
- • Data Matrix codes
Document Annotations
- • Handwritten signatures
- • Redlines and markup
- • Sticky notes and comments
- • Stamps and seals
Precision Data: All extracted elements include bounding box coordinates, confidence scores, and metadata for precise positioning and validation.
Cardinal is built with enterprise security and compliance in mind:
- • Zero Data Retention: Documents are processed and immediately deleted
- • HIPAA Compliance: Business Associate Agreements available
- • SOC 2 Type II: Certified for security, availability, and confidentiality
- • On-premise deployment: Keep data within your infrastructure
- • End-to-end encryption: Data encrypted in transit and at rest
On-Premise Deployment
Need to keep your data on-premise? Cardinal offers flexible deployment options for enterprise customers with strict security and compliance requirements.