Guides
Quick Start
Set up your first end-to-end OCR flow in Numora.
1. Prepare Your Workspace
After signing in, open the file workspace and confirm your team has upload permissions.
2. Upload Sample Documents
Upload a small sample set first (for example 10 to 20 files) so you can validate extraction quality before scaling.
3. Run OCR Extraction
Trigger extraction and wait for processing to finish.
For best results:
- Use readable scans.
- Avoid severe skew and rotation.
- Keep document categories separated when possible.
4. Review and Confirm Fields
Check extracted values, especially low-confidence fields and business-critical fields such as totals, dates, and IDs.
5. Export or Integrate
After confirmation, export structured results or send them to your downstream workflow.
6. Move to Production
Before production rollout, define:
- Review policy (what must be human-checked).
- Error handling and retry rules.
- Retention and access controls.
See Data and Security for operational guidance.
7. Plan Your API Integration
If you want to embed Numora into another product or connect it to operational systems, review the public API developer preview docs next:
- API Authentication and First Request
- Authentication and First CRM Delivery
- Public API Overview
- Document API
- n8n Outbound Delivery
- Polling Reviewed Documents
- Delivery Status Callback
- Document Lifecycle
- Review and Approval
- Mappings and Destinations
- Workflow Runs
- Error Codes and Limits
- Webhook Delivery
- Public API Conventions
- Changelog