Cradl AI

Cradl AI positions itself as the document AI layer for automation workflows. It processes high volumes of complex, multi-page documents, splits and classifies them, and performs line-level extraction for any layout and language. Predictions receive confidence scores, cross-model checks, and customizable business rules before results reach downstream systems. Exceptions can be routed to business users through a dedicated UI, with corrections captured and source evidence shown for validation. The platform connects through REST APIs and webhooks and can be added to Power Automate, Zapier, or n8n workflows. The website's FAQ states that data retention, access, and deletion are configurable, that customer data is not used to train shared models, and that the platform supports low-code teams as well as API-first developers. Example customers named on the page include Equinor, Posten Bring Logistics, SpareBank 1 SR-Bank, and Yara International.

Key features

• Split and classify documents
• Line-level extraction and classification
• Extraction from complex, multi-page documents in any layout and language
• Confidence scores for every field
• Cross-model checks
• Fully customizable business rules, with code when needed
• Anti-hallucination check
• Dedicated UI for business users to review exceptions
• Learns from corrections
• Source evidence showing where each prediction was extracted from
• REST API and webhook integrations
• Integration with Power Automate, Zapier, and n8n
• Configurable data retention, access, and deletion settings
• Low-code support alongside API-first integration paths

Best suited for

• Automation teams at mid-sized and enterprise companies
• Document-heavy processes automated through Microsoft Power Automate, n8n, Zapier, or APIs
• Invoice data extraction
• PDF parsing and PDF to Excel conversion
• Processing passports, bills of lading, purchase orders, invoices, and inventory lists

Pricing

Not stated on the website