The AI-900 exam has been retired
It was retired on June 30, 2026. It was replaced by AI-901 (Microsoft Azure AI Fundamentals). Go to the AI-901 study guide →
2.1.3. Document Processing Workloads
Document processing workloads handle structured and semi-structured documents like invoices, receipts, forms, and IDs. These workloads combine multiple AI capabilities—primarily OCR (to read text) with intelligent extraction (to understand what the text means).
Key characteristics:
- Extracts structured information from business documents
- Understands document layout and field relationships
- Can process forms, invoices, receipts, and identity documents
- Goes beyond simple OCR to understand document semantics
- Returns structured data (key-value pairs, tables) not just raw text
How document processing differs from simple OCR:
| Capability | OCR | Document Processing |
|---|---|---|
| Output | Raw text | Structured fields (invoice_number, total, vendor) |
| Understanding | Just reads characters | Understands document type and field meanings |
| Tables | May struggle | Extracts table data with headers |
| Use case | "What text is here?" | "What's the total on this invoice?" |
Common scenarios:
- Invoice processing: Extract vendor, amounts, line items, dates
- Receipt scanning: Capture merchant, items, totals for expense reports
- Form data extraction: Pull field values from filled forms
- ID verification: Extract name, DOB, document number from identity documents
- Contract analysis: Extract parties, dates, key terms from legal documents
Azure service: Azure AI Document Intelligence (formerly Form Recognizer) handles document processing workloads with pre-built models for common document types and custom model training for specific forms.
⚠️ Exam Tip: If a question mentions extracting STRUCTURED DATA from documents (invoices, forms, receipts), think Document Processing. If it just asks about reading text from any image, think OCR.
