Azure
AI & ML
intermediate
Document Intelligence Pipeline — Azure Architecture Template
Uploaded documents trigger an extraction function that combines Form Recognizer and OpenAI summarization, with human review workflow.
Why this architecture works
- Event-driven extraction starts the moment a document lands — no polling, no batch delay
- Form Recognizer handles structured fields while OpenAI summarizes free text; each model does what it is best at
- Low-confidence extractions route to a Logic App human-review step instead of silently entering the system
- Extracted data in Cosmos DB keeps the original blob immutable as the audit source of truth
- AI service keys live in Key Vault and per-document processing is traced in Application Insights
What's inside (9 resources)
Doc Inbox
storage-account
Blob Events
event-grid-topic
Form Recognizer
applied-ai
Extractor
function-app
Summarizer
openai
Extracted Data
cosmos-db
Review Workflow
logic-app
Secrets
key-vault
APM
application-insights
From template to running infrastructure
- Open this template in the CloudForge visual designer (free account, no credit card).
- Customize resources, names, and connections on the drag-and-drop canvas — or ask Vani, the AI architect, to adapt it.
- Generate production-ready Terraform, ARM, or Bicep in one click.
- Review the plan diff and security scan, then deploy with human approval.