Digitoo.ai
Processes

Extracting data

In this case a client wants to just extract data from a document.

Prerequisites:

  1. Active account at app.digitoo.ai and authentication token (how to get auth token)
  2. Created workspace within organization (see more)
  3. Queue ID (how to get Queue ID or Queue ID from user response)

Upload a document

  1. Call upload API

    Import of document is initiated via POST call to /api/v2/queues/:queueId/upload

    Replace :queueId with Queue ID eg. 460c2ced-87e2-4fe3-b8ac-55b4d955382c

    TerminalCode
     
    curl
--location 'https://api.digitoo.ai/api/v2/queues/:queueId/upload' \
--header 'Authorization: Bearer :authToken' \
--form 'file=@"/Users/Downloads/invoice.pdf"'
--form 'notify_webhook_url="https://webhook.url"'

  2. Add webhook url if needed

    You are able to specify notify_webhook_url as optional parameter. We will use this URL to notify you when document's status is changed. The URL must handle POST request and will have following body:

    JSONCode
     
    {
  "document_id": "id",
  "queue_id": "queue_id",
  "status": "ready-to-export"
}

  3. Webhook call

    After document is upload, you shall receive document ID which can be used to call further methods on the document eg. to GET document annotations by calling /api/v2/documents/:documentId/annotations-values

Invalid invoice detection

In case when extraction service detects the uploaded document is not a valid invoice (eg. it's an email attachment of other type such as a company logo) – it skips the annotation extraction process and the document is set to extraction-rejected status. These documents can then be found under Spam folder in Digitoo App.

If the AI detects a false positive, the document can be send to extraction again using this endpoint: #api-v2-documents-enqueue-to-extraction

In case you do not want this feature, you can disable via API: #api-v2-queues-use-invalid-invoice-detection

Last modified on
OpenAPISupported file types