Automation Anywhere provides three AI-powered capabilities for document processing and knowledge extraction: Document Automation, AI Skills, and Automation Anywhere Enterprise Knowledge. Use this topic to select the right capability for your automation scenario based on input type, output requirements, and accuracy needs.

Each capability is optimized for different use cases, from structured data extraction to flexible document understanding and conversational question answering across enterprise content.

How to use this guide

Select the right capability based on:

  • The type of input: structured forms, unstructured documents, or knowledge repositories
  • The type of output required: exact fields, summaries, or answers
  • The level of accuracy, flexibility, and control required

Compare capabilities


Choose Right AI capability
Capability Best for Input type Output type Strengths Considerations
Document Automation Extracting data from business documents (fixed forms, semistructured or unstructured documents) Documents (PDF, scanned images, forms) Structured data (fields, tables) High accuracy for known formats; deterministic extraction; built-in training; guaranteed output schema; adaptive learning Requires setup and training; not designed for one-off Q&A, summarization, or translation tasks
AI Skills (Multimodal models) Flexible document understanding and content generation Text and files (images, PDFs, documents) Generated text, summaries, structured or semistructured output Handles unstructured data; supports multiple file types; no training required Output is generative; requires prompt design; less deterministic than tuned extraction tooling
Automation Anywhere Enterprise Knowledge (RAG) and Automation Co-Pilot Question answering and conversational assistance using enterprise content Indexed documents, enterprise knowledge bases, natural language queries Conversational responses, contextual answers, suggestions Easy-to-use conversational interface; grounded enterprise answers; works across large datasets Requires indexing and grounding setup; not designed for precise field extraction

When to use each capability

Document Automation

Use Document Automation when:

  • You need to process a business document (text PDFs, image PDFs, scans, or multipage documents)
  • You need structured output
  • You need validation rules to verify extracted values
  • Extraction accuracy is critical
  • You need output schema guarantees for downstream automation
  • You need a specialized UI for reviewing or correcting extracted data
AI Skills (Multimodal)

Use AI Skills when:

  • You need summaries, insights, or generated content
  • You need to combine text and files in prompts
  • Documents are unstructured or vary in format
  • You need flexible, prompt-driven interactions
Automation Anywhere Enterprise Knowledge (RAG) and Automation Co-Pilot

Use Automation Anywhere Enterprise Knowledge or Automation Co-Pilot when:

  • You need conversational answers across enterprise content
  • You need contextual answers grounded in enterprise knowledge
  • Users prefer chat-based interactions and workflow guidance

Choose the right tool

Need exact field extraction
Use Document Automation. See Document Automation.
Need summaries or insights from files
Use AI Skills. See Skills.
Need answers from enterprise knowledge
Use Enterprise Knowledge. See Automation Anywhere Enterprise Knowledge.
Need conversational help or workflow guidance
Use Automation Co-Pilot+. See Automation Co-Pilot.