Skip to main content
Follow these steps to create a workflow, build a simple workflow, and extract structured data from a document.

​
Prerequisites

Ingestly is currently in private beta. If you don’t have an account yet, request access at ingestly.ai/beta. Once your invite is approved you’ll receive an email with a link to complete signup.
1

Sign up and log in

Go to ingestly.ai and create your account. After signing in, you land on the Dashboard. Click Workflows in the sidebar to continue.
2

Create your first workflow

Click New Workflow and enter a name (for example, “Invoice Extraction”). A workflow is the named process that holds your configuration, uploaded documents, and processing runs.
3

Build the workflow

Navigate to the Workflow tab inside your new workflow. The visual editor opens with an empty canvas.Add these three nodes to create a basic extraction workflow:
  1. Upload trigger: accepts document uploads
  2. Extract action: uses AI to extract structured data from the document. Configure a schema with the fields you want to extract (for example: vendor_name, invoice_number, total_amount).
  3. Download output: makes the extracted data available as a structured JSON file you can download once the run completes
Connect the nodes in order: Upload Trigger → Extract Action → Download Output.Click Save to save your workflow, then Activate to enable it.
4

Upload a document

Inside your workflow, click Upload. In the upload dialog, drag and drop a document or click to browse. Ingestly supports PDF, JPEG, PNG, TIFF, GIF, WebP, and CSV, plus Word, Excel, PowerPoint, OpenDocument, and Rich Text documents. See supported formats for the exact extensions.Once uploaded, a run starts automatically. You can watch the progress in real time.
5

View results

Click the Runs tab inside your workflow. Click on a run row to open its detail page, which shows each step in your workflow, its status, and the extracted data.Once the run completes, use the Download output step on the run detail page to download the structured result as a JSON file.

​
Next steps

Build your first workflow

Learn more about the visual workflow editor and all available node types

Core concepts

Understand workflows, runs, and review tasks in depth

Schema design

Design effective extraction schemas for your documents

API access

Trigger workflows programmatically via the HTTP API