Joe Filcik announces the general availability of Azure Content Understanding in Microsoft Foundry Tools, detailing enhancements in model deployment, document analysis, generative AI integration, and practical customer applications across industries.

Azure Content Understanding Now Generally Available in Foundry Tools

Azure Content Understanding has reached general availability (GA) within Microsoft Foundry Tools, introducing industry-grade, production-ready features for intelligent document analysis and retrieval-augmented generation (RAG) solutions.

What’s New

  • Direct Connection to Foundry Models:
    • Deploy Content Understanding with any Microsoft Foundry model.
    • Configure for quality, cost, and latency to match your use cases.
    • Supports pay-as-you-go and Provisioned Throughput Units (PTUs) with expanded global availability.
  • Integration with Foundry IQ and Azure AI Search:
    • Seamlessly connect content extraction with enterprise search and agentic workflows, streamlining ingestion and metadata enrichment.
  • Model Flexibility and Architecture:
    • Combine specialized models (for OCR, layout, and transcription) with generative AI (for field extraction, segmentation, figure analysis).
    • Enable predictable costs and high-quality outputs for various applications.
  • Expanded Region Availability:
    • Now available in 12 regions, offering improved data residency and lower latencies.

Prebuilt and Domain-Specific Analyzers

  • Prebuilt Analyzer Library:
    • Ready-to-use analyzers for top use cases including RAG ingestion, finance, procurement, and more, extending Azure Document Intelligence with foundation models.
  • RAG-Specific Tools:
    • Extract tables, paragraphs, sections, and figures for richer structured outputs essential for retrieval-augmented generation.
  • Multimodal Support:
    • Prebuilt analyzers for video, audio, and images enable structured outputs and context-aware data for agentic or search solutions.
  • Layout-Aware Enrichment:
    • Extracts structured information like tables (even spanning pages), figures, diagrams, and charts, outputting formats like Chart.js and Mermaid.js for downstream use.
  • Domain-Specific Accelerators:
    • Analyzers for finance (statements, tax forms), contracts & procurement, ID verification, mortgage and lending, and more for fast, accurate field extraction.

Intelligent Document Processing and Workflow Upgrades

  • Segmentation and Classification:
    • Automatically split complex documents (e.g., visa applications) into their components, routing each part to appropriate analyzers.
  • Granular Control and Cost Reduction:
    • Enhanced support for confidence scores, grounding, complex schemas, and field-level extraction choices, now including mini and nano models for cost-efficient processing.
  • Operational Improvements:
    • Automatic method selection (extract/generate), support for complex object types, and consistency across modalities.

Integration with Agents and Multi-Agent Systems

  • Standardized file representation and schema for developers building agent workflows.
  • Upcoming Content Understanding MCP server for advanced agentic integrations.

Customer Success Stories

  • KPMG: Uses Content Understanding within KPMG Clara AI to transform unstructured audit data into structured, actionable insights, streamlining complex audit engagements.

  • DataSnipper: Empowers finance teams to extract and validate data within Excel using Azure Content Understanding, linking extracted values to original sources for full auditability and efficiency.

  • SJR: Embeds Content Understanding in their GX Manager, achieving personalization at scale and paradigm shifts in digital marketing with structured, AI-powered content.

Getting Started

Conclusion

Azure Content Understanding, now generally available, provides a flexible and robust foundation for building intelligent document and content processing solutions with native integration across the Microsoft ecosystem. This release brings together advanced AI, multimodal support, prebuilt and domain-specific analyzers, and improved developer tooling, enabling organizations to translate content complexity into actionable insights at enterprise scale.

This post appeared first on “Microsoft AI Foundry Blog”. Read the entire article here