Skip to content

Overview

Turn documents into structured data, from your first template to a production backend integration. These guides cover building and testing extraction in the app, then wiring it into your own systems through the API.

Most teams set up extraction in this order:

  1. Create a template for one document type, with the fields you want to extract.
  2. Test it in Playground or Quick Run until the output looks right.
  3. Decide how files get processed going forward: pick them up automatically with workflows (no code), or call the API yourself (synchronous or async).
  4. If you’ll call the API, create an API key.
  5. Optionally add webhooks or document review if downstream systems need automation or human approval.
GoalRecommended pathRead next
Try extraction on one document by handTemplate plus Playground or Quick RunTemplates, Filters & Workflows
Process files automatically, no codeWorkflowTemplates, Filters & Workflows
Set up API accessAPI keyAuthentication & Errors
Extract one file and get the result nowSynchronous requestSynchronous Extraction
Run extraction from your backend in productionAsync jobAsynchronous Extraction
Route different document types automaticallyFilterTemplates, Filters & Workflows
Get notified when async work finishesAsync job plus webhookWebhooks
Hold low-confidence results for reviewAsync job or workflow with document reviewAsynchronous Extraction
See all API endpointsEndpoint referenceEndpoint Reference