easybits logo

n8n Integration Guide

Every pipeline you created on easybits' Extractor platform exposes individual API-endpoint. You can send your data there via regular HTTP-request. Below is the step-by-step guide how to add a data extration step to your n8n-workflow.

Step 1: Create a new workflow

Create a new workflow or use our example.

Step 2: Add a "HTTP Request" Node

Select HTTP Request node in the list of nodes and add it to your workflow.

Select HTTP Request Node

Step 3: Configure the Node

In the HTTP node settings, configure the following values. Take the necessary information from the pipeline details page in the extractor app.

PropertyValue
MethodPOST
URLhttps://extractor.easybits.tech/api/pipelines/[YOUR_PIPELINE_ID]
AuthenticationPredefined Credential Type
Credential TypeBearer Auth
Bearer AuthAdd Your Credential

Click “+ Create new credential” to add your API key. Fill in the credential details as follows:

  • Bearer Token: [YOUR_PIPELINE_API_KEY_HERE]

Click “Save” to store the credential. Continue configuring the HTTP node:

PropertyValue
Send BodyON
Body Content TypeJSON
Specify BodyUsing JSON

After completing the configuration, your HTTP node should look like this.

HTTP Request Node settings

Step 4: Define the Request Format

Our API supports the following file types:

  • Image files (.png, .jpeg)
  • PDF documents (.pdf)

There are two ways of sending data to the pipeline. If your file is available via URL in some cloud storage, you can simply pass the link as a string.

{
  "files": [
    "https://example.com/document1.pdf"
  ]
}

If you do not store file and want to process them on the fly, you will need to convert them to Base64 Data URL.

{
  "files": [
    "data:<MIME TYPE>;base64,<BASE64-ENCODED FILE CONTENT>"
  ]
}

Sending multiple documents

You can send several files in a payload, but keep in mind that all of them will be treated as one document. Uploading multiple different documents can lead to inaccurate extraction and unreliable results.

  • Good: multiple images, each image is a single page of the same document.
  • Bad: multiple images, each image is a separate document you want to extract information from.

Example cURL Request

If you're not using n8n, you can still integrate your pipeline with any kind of app via regular API request.

curl -X POST https://extractor.easybits.tech/api/pipelines/YOUR_PIPELINE_ID \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "files": [
      "https://example.com/invoice-page1.pdf"
    ]
  }'

Response Format

The API will returns the extracted data in JSON format, following the response structure (schema) defined in your pipeline mapping.

{
  "data": {
    "invoice_no": "extracted value",
  }
}