Intelligent Document Understanding

osop_version: "1.0"
id: "document-understanding"
name: "Intelligent Document Understanding"
description: "OCR extraction, entity recognition, relationship mapping, and knowledge graph construction."

nodes:
  - id: "ocr_extract"
    type: "agent"
    subtype: "worker"
    name: "OCR Extraction Agent"
    description: "Convert scanned PDFs and images to structured text with layout preservation."

  - id: "entity_recognition"
    type: "agent"
    subtype: "llm"
    name: "Entity Recognition Agent"
    description: "Identify people, organizations, dates, amounts, and domain-specific entities."

  - id: "relationship_map"
    type: "agent"
    subtype: "llm"
    name: "Relationship Mapping Agent"
    description: "Detect relationships between entities: ownership, employment, contractual obligations."

  - id: "knowledge_graph"
    type: "agent"
    subtype: "llm"
    name: "Knowledge Graph Agent"
    description: "Merge extracted relationships into a persistent knowledge graph with deduplication."

  - id: "structured_output"
    type: "api"
    name: "Structured Output"
    description: "Export JSON-LD, CSV, or graph database triples for downstream consumption."

edges:
  - from: "ocr_extract"
    to: "entity_recognition"
    mode: "sequential"
  - from: "entity_recognition"
    to: "relationship_map"
    mode: "sequential"
  - from: "relationship_map"
    to: "knowledge_graph"
    mode: "sequential"
  - from: "knowledge_graph"
    to: "structured_output"
    mode: "sequential"
  - from: "entity_recognition"
    to: "ocr_extract"
    mode: "fallback"
    label: "Low confidence entities, re-OCR with enhanced preprocessing"