Introduction

Users can integrate AI Workflows with ElectroNeek's own products, not only those within the AI Agent Management module, but also the Document Intelligence and RPA Bot Orchestration ones.

These integrations, built into the ElectroNeek Piece, leverage the built-in APIs on these modules to integrate RPA Bots and Document Streams into your workflows.

Creating ElectroNeek Connections

In order to leverage the ElectroNeek integrations in your flows, you need to create Connections for Orchestrator (for RPA-related Actions) and IntelliDocs (for document processing). The Portal Connection used by Public Forms is provided automatically.

In the following sections, you'll find details on how these connections work and how to set them up.

Portal Connection

If you navigate to the Connections page, you will find a pre-existing one called ElectroNeek Portal Connection. This is pre-configured with a connection to your environment that can be leveraged by some features, including Public Forms.

❗️

Warning

Do not delete or edit the ElectroNeek Portal Connection. It can be manually re-created only by an ElectroNeek team member.

Orchestrator Connection

📘

Orchestrator REST API required

In order to create this Connection, the Orchestrator REST API feature must be enabled in your organization. You will not be able to generate an API Token if this feature is not enabled.

For RPA-related Actions, a Connection to Orchestrator has to be created. The fields required for that Connection are:

• Connection Name — can be anything, but we recommend including the word "Orchestrator" to differentiate it from other ElectroNeek connections.
• Bearer Token — paste the Token from Orchestrator SaaS > Integrations > API.
• API URL — https://api.{region}.electroneek.com/v1/orchestrator (replace {region} with your region code: latam, na, emea, or apac).

IntelliDocs Connection

For IntelliDocs Actions, each Document Stream must have its own Connection. Here's how to set it up:

• Connection Name — can be anything, but we recommend a scheme like Doc Stream: {document type} or IntelliDocs: {document type}.
• Bearer Token — paste the Document Stream's API Key.
• API URL — https://api.electroneek.com/v1/intellidocs (the same for all regions).

📘

Keep in mind

These Connections must be created manually for now. Automatic provisioning may be introduced in a future release.

Triggers

Triggers start an AI Workflow when an event occurs in ElectroNeek.

Public Form Submitted

Triggers when a public form is submitted.

🚧

Connection Required

This Trigger relies on the "ElectroNeek Portal Connection" Connection that you see in the "Connections" page. This Connection is added by default. Do not remove or delete it.

Testing this trigger

1. Open the Public Form in a new browser window (use the link from your ElectroNeek Portal > Forms page.)
2. Fill in the Form fields and submit it.
3. Return to the AI Workflow and click Test Trigger to capture the submission as sample data.

The captured fields will be available as variables in subsequent flow steps through the Data Selector.

⚠️

Warning

The Form cannot be linked to a Legacy Process to work with this trigger. If it is, the trigger will not work. Make sure to unlink it first, or create a new Form.

Form

Select a Form from the list. Only Forms that have an active Public link will be listed.

Output

When a form is submitted, the Trigger outputs a JSON object containing the submission metadata and the submitted values. Text fields are returned as strings, and file-upload fields are returned as an array of file objects.

The example below is from a form with two text fields (question1, question2) and one file-upload field (question3). The base64 content is truncated here for readability:

{
  "submissionId": "6a285fcc5e0412359d39be1a",
  "formId": "69e644d63a056fd175029a6a",
  "result": {
    "question1": "John Doe",
    "question2": "[email protected]",
    "question3": [
      {
        "name": "screenshot.png",
        "type": "image/png",
        "content": "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAA…"
      }
    ]
  },
  "parsedResult": {
    "question1": "John Doe",
    "question2": "[email protected]",
    "question3": [
      {
        "name": "screenshot.png",
        "type": "image/png",
        "content": "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAA…"
      }
    ]
  },
  "submittedAt": "2026-06-09T18:47:40.452Z",
  "createdAt": "2026-06-09T18:47:40.453Z"
}

The keys inside result and parsedResult (question1, question2, …) mirror your form's fields, so they vary from form to form. Reference each value using the field name from your own form.

File-upload fields return an array of objects (one per uploaded file, base64-encoded), each with:

Actions

Actions are the steps an AI Workflow performs against ElectroNeek's products.

Run RPA Workflow

Launch an ElectroNeek RPA workflow on your Orchestrator and optionally specify a Bot Runner and input payload.

Connection

Select the Orchestrator Connection this Action should use. If you haven't created one yet, see Orchestrator Connection above for the required fields and setup steps.

Workflow

Select the Workflow to launch. Only workflows from the connected Orchestrator will be listed.

Bot Runner

Optionally select a specific Bot Runner. If left empty, the first available Bot Runner is used.

Input Values Mode

How to supply the workflow input payload. Use Advanced (the default) to provide a raw JSON object via the JSON Payload field — this is the recommended mode for Agent tools. Use Simple only when building the payload manually through the UI variable list.

JSON Payload

The JSON object sent as the workflow input payload (used when Input Values Mode is Advanced). Provide a plain object containing the variables the bot expects, for example:

{ "run_type": "GetSpecificFile", "file_path": "C:\\Files\\image.png" }

Leave it empty if the workflow requires no input.

Variables

Add input variables manually through the UI to pass to the workflow. These are the variables passed as input when Input Values Mode is set to Simple.

Continue on Failure

Enable this option to skip this step and continue the flow normally if it fails.

Retry on Failure

Automatically retry up to four attempts when failed.

Output

This Action launches the workflow through the Orchestrator Launch Workflow API method and returns its response.

On success, it returns the launch ID. Pass it to the Wait for RPA Workflow Action (as the Workflow Launch ID) to track the run to completion:

{
  "launch_id": "6a2c409e920460d4583b6536"
}

If the launch fails — for example, when no Bot Runner is assigned to the workflow — it returns the error response instead:

{
  "response": {
    "status": 403,
    "body": {
      "message": "Bot runner is not assigned to workflow",
      "error": "400:WORKFLOW_PUBLIC_API:BOT_RUNNER_NOT_ASSIGNED_TO_WORKFLOW",
      "statusCode": 403,
      "timestamp": "2026-06-12T17:20:11.312Z",
      "path": "/public/workflow/633d462feb200ddf4e7dac42/launch"
    }
  },
  "request": {}
}

Wait for RPA Workflow

Pauses the flow until a launched RPA workflow completes (successfully or with error), then returns the full launch details.

Connection

Select the Orchestrator Connection this Action should use. If you haven't created one yet, see Orchestrator Connection above for the required fields and setup steps.

Workflow Launch ID

The Launch ID returned by the Run RPA Workflow Action.

Timeout (seconds)

Maximum time to wait for the workflow to complete. Leave empty for the default timeout of 24 hours.

Polling Interval (seconds)

How often to check the workflow status. Defaults to 30 seconds.

Continue on Failure

Enable this option to skip this step and continue the flow normally if it fails.

Output

This Action checks the launch status on each polling interval until the workflow reaches a final state, then returns the full launch details. These details are the same as those provided by the Get Launch Info API method — but, unlike a single Get Launch Info call (which returns the current status immediately), this Action waits for the run to finish before returning.

The status field indicates the outcome and is one of in-progress, completed-successfully, or completed-with-error.

{
  "id": "6a2c426b920460d4583c0b96",
  "logsEnabled": true,
  "consoleLogsEnabled": true,
  "runner_id": "69efaf67a886d40e1fd4804f",
  "started_at": "2026-06-12T17:31:25.178Z",
  "status": "completed-successfully",
  "type": "api",
  "workflow_id": "633d462feb200ddf4e7dac42",
  "account_email": "[email protected]",
  "completed_at": "2026-06-12T17:31:35.095Z"
}

📘

Output data

If the workflow was built in Studio Pro to return Output variables, the launch details may also include an output_data object containing those values.

Extract Data with IntelliDocs

Send a document to IntelliDocs for recognition and extract structured data from it. Uses the Document Stream identified by your API key.

Connection

Select the IntelliDocs Connection this Action should use. If you haven't created one yet, see IntelliDocs Connection above for the required fields and setup steps.

Document

A file input — a URL or a file passed from a previous step. The document file to send for recognition. Supported formats: PNG, JPEG, GIF, BMP, WEBP, PDF. Maximum file size is 20 MB.

Continue on Failure

Enable this option to skip this step and continue the flow normally if it fails.

Retry on Failure

Automatically retry up to four attempts when failed.

Output

This Action sends the document through the IntelliDocs Send Document to Recognition method and returns the extracted data. The contents of fields and ocr are fully determined by your Document Stream — the values below are only an illustration (from an invoice stream):

{
  "outputVersion": "v2",
  "fields": {
    "Invoice Number": "123456A",
    "Total Amount": 4520,
    "Items": [ { "Item Description": "500 Lb Anvil", "Item Quantity": 1, "Item Price": 4520 } ]
  },
  "ocr": "SAMPLE INVOICE\nBilled To:\nClient Name\n...",
  "tableOcr": "",
  "usage": {
    "completionTokens": 63,
    "promptTokens": 1165,
    "totalTokens": 1228
  },
  "recognizedPages": 1,
  "errors": []
}