Update a RAG Collection from a Process

  • 1 minute(s) read

Add Files to a RAG Collection from a Process

Follow these steps to add files to a RAG collection from a process.

Prepare Process Variables

From your process, ensure you have the following variables:

  • File ID – The variable name of the uploaded document (e.g., file_id).

  • File Name – The name to assign to the file when saved in the collection (e.g., file_name).

Create a REST Data Connector

  1. After creating a REST Data Connector, set up a REST Data Connector using the Bearer Token authentication method.

  2. Provide a valid Bearer Token to authenticate the connection with your ProcessMaker instance.

  3. Create a Resource in the Data Connector that uses the following API endpoint for collections:

bashCopiarEditar/api/1.0/collections/{collection_id}/records

Replace {collection_id} with the ID of your target RAG Collection.

  1. In the Body of the Resource, use the following JSON structure:

{
  "data": {
    "file": "{{file_id}}",
    "name": "{{file_name}}",
    "uploads": [
      {
        "id": "{{file_id}}",
        "name": "{{file_name}}"
      }
    ],
    "source_type": "file"
  }
}

Add a Data Connector Task

Insert a Data Connector Task in your process, ensuring it comes after the file is uploaded and the file-related variables are populated.

Configure Outbound Parameters

Map the process variables to the connector resource in the task's Outbound Configuration:

  • file_id → {{file_id}}

  • file_name → {{file_name}}

Result

Each time a case runs, the Data Connector will automatically insert the uploaded file as a new record in the RAG Collection, triggering the appropriate analysis.

To send the file_name field to the Data Connector, make sure to enter the full file name including the file extension (e.g., document.pdf, invoice.docx).

Related articles