AkaXtract
AkaXtract turns document content into structured data. You upload a document, AI classifies it and extracts the data, you review the result, and you export it or send it to another system.
How It Works
- Set up a Pipeline that defines what to extract.
- Test and publish it.
- Upload documents manually, or receive them through ScaleConnect or the API.
- Review results that need a person to confirm them.
- Export the data, or let a Webhook deliver it.
Key Concepts
| Term | Meaning |
|---|---|
| Pipeline | The processing rules: AI model, classification instructions, extraction instructions for each document type, and approval rules |
| Document | A source file and the review status of its result |
| Job | One processing run, with its own ID, Pipeline version, status, and errors |
| Human Review | A reviewer compares the extracted data with the source document and confirms it |
| Webhook | An automatic notification sent to a receiving system |
Choose Your Path
| I am… | Start with |
|---|---|
| New to AkaXtract | Quick Start |
| Setting up document processing | Create a Pipeline, then Test and Publish |
| Learning how AkaXtract is configured and how to improve accuracy | Learning Method |
| Uploading or reviewing documents | Upload and Track, then Review and Export |
| Connecting another system | Webhooks and API Integration |
| Managing users and quota | Users and Roles and Quota and Retention |
| Stuck on an error | Troubleshooting |
note
Menu and button names follow the English interface. The interface is also available in Vietnamese, Japanese, and Traditional Chinese. The menus you see depend on your permissions.