document-intelligence

What problem does it solve?

Automate structured data extraction and document parsing from PDFs and other document formats using Snowflake Cortex AI, transforming unstructured content into usable data for downstream analytics, workflows, and decision-making.

Core Features & Use Cases

• End-to-end document processing with AI_EXTRACT (structured fields/tables), AI_PARSE_DOCUMENT (full content with layout), and AI_COMPLETE (visual content analysis) for invoices, contracts, reports, charts, and blueprints.
• Batch and pipeline readiness: build processing pipelines, handle large documents via page filters, layouts, and optional page-splitting strategies.
• Multi-format support: PDFs, images, Office documents, HTML/TXT, and more, with capabilities for data extraction, layout preservation, and image analysis.
• Real-world use cases: extract invoice data into structured fields, parse research documents for RAG-ready content, and analyze charts/blueprints to produce structured outputs.

Quick Start

Use the Document Intelligence skill to extract invoice data from a file located in your Snowflake stage, for example invoice.pdf. Then apply AI_EXTRACT with a field mapping to capture invoice_number, vendor, and total_amount, and store the results as needed.