Extract Document Data

Prerequisites

Input options

Supported languages for generative models

Models used by the activity

The Extract Document Data activity uses the following:

• Pre-trained specialized models available out of the box, based on DocPath.
• Custom pre-trained models deployed in Document Understanding modern and classic projects.
• Generative extraction models.

Designer panel

• Input - Requires you to specify the file itself, or Document Data, in case you have used other Document Understanding Activities before in your workflow, (for example, Classify Document).
  Important: The maximum numbers of pages a file can have is 500. Files exceeding this limit fail to extract.
• Project - Requires you to select your Document Understanding project from the dropdown list. The available options are:
  • Predefined – Classic project type that uses pre-trained specialized models recommended for standard scenarios.

    For more information on the charging logic for classic project, visit Metering and charging logic.

  • Generative Predefined – Modern project type that uses pre-trained generative models accepting instructions as input for extraction of document data.

    For more information on the charging logic for modern projects, visit Metering and charging logic.

  • Existing projects from the tenant and folder you are connected to.
  • You can create a custom project by going to Document Understanding.

    For more information, visit Introduction for building models.

  Note: If you have created more than 500 projects on your tenant and use the Extract Document Data activity, UiPath Studio or Studio Web will not display any projects beyond the initial 500. Therefore, those projects cannot be used.

• Extractor - After you select a project, you can also select an extractor that you want to use.
  • For the Predefined project, you have two choices:
    • Select a pre-trained model. Visit Out-of-the-box models for a list of pre-trained models that you can use.
      Note: The Extract Document Data activity extracts the information for the fields available on the document type for the selected extractor (regardless of the actual type of the document). This is not applicable for generative models.
    • Select the Generative extractor.
      Note: The information sent to the Generative Extractor goes to an LLM Model instance. This instance isn't publicly available, doesn't store the data sent, and doesn't use it for training purposes.
      Important:

      This feature is currently part of an audit process and is not to be considered part of the FedRAMP Authorization until the review is finalized. See here the full list of features currently under review.

  • For the Generative Predefined project, you have three choices for extraction, tailored to a specific document layout:
    • Long Document Simple Layout Extractor – Recommended for long form documents with mostly text and headings. For example, you can use the Long Document Simple Layout Extractor on documents such as lease agreements, master service agreements, or other similar documents.
    • Long Document Complex Layout Extractor – Recommended for long form documents that include elements such as images, handwriting, form controls, floating callout boxes, or other complex layout types. For example, you can use the Long Document Complex Layout Extractor on documents such as insurance policies, or other similar documents.
    • Short Document Complex Layout Extractor – Recommended for short documents that include elements such as images, handwriting, form control, floating callout boxes, or other complex layout types. For example, you can use the Short Document Complex Layout Extractor on documents such as government IDs, healthcare intake forms, or other similar documents.
  • Use Classification Result: If the Generate Data Type property is set to false, you can opt for the Use Classification Result option. This option automatically uses a recommended extractor based on the document type resulted from the Classify Document activity.

    If multiple extractors can work with that document type, the activity returns an error. In this scenario, you must manually select your preferred extractor.

• Document Type details - This field appears if you choose the option Generative. Prompt to identify the fields to be extracted, provided as key-value pairs, where the key represents the name of the field and the value a description for it, helping the extractor identify the corresponding value. Select the field, and you will get a prompt with the following options, provided as pairs:
  • Field name - Requires you to input the field name to be extracted (Ex. Due date) (30-character limit)
  • Instruction - Requires you to provide instructions about what information should be extracted for the corresponding field.. The maximum number of characters allowed is 1000. The response, extraction result, also called Completion, has a word limit of 700. This is limited to 700 words. This means that you can't extract more than 700 words from a single prompt. If your extraction requirements exceed this limit, you can divide the document into multiple pages, process them individually, and then merge the results afterwards.
  Tip: For good practices on how to use generative prompts, check the Generative extractor - Good practices page.
• Version or Tag - Use this property when using an existing Document Understanding modern project. Select the tag that corresponds to the project version from which you want to process data. For instance, if you choose the Production tag assigned to Version 3, the activity processes data from Version 3 of your project in the production environment.

  The default value for Version is Staging. If the Staging tag doesn't exist in your selected project, then the default value is Production.

  For more information about versions, visit Publishing models.

• Document Type - When you choose a tag from the Version field, the activity automatically selects the first deployed document type from the relevant version of your chosen project. Moreover, the activity shows the extraction fields related to your chosen document type.

Properties panel

• Timeout (seconds) - Maximum execution time (in seconds) for the call to the generative model. If the operation exceeds this timeout, it is automatically terminated to prevent delays or hangs. This property is only displayed if the Generative Extractor is selected as an extractor.
• Auto-validation - Use this option to enable automatic validation, a capability that helps validate the results obtained for data extraction against a Generative model. The default value for the Auto-validation field is False.
  • Confidence threshold - This field becomes visible once you enable Auto-validation. Extraction results falling below the threshold are compared to the generative extraction model. If they match, the system adjusts the extraction confidence to meet the threshold value. Possible threshold values range from 0 to 100.

    If the value is set to 0, no validation is applied. However, if you set a specific value (from 0 to 100), the system checks all extraction results below this value. For example, if you set a confidence threshold of 80%, the system will apply the generative validation for fields with confidence below 80%.

    Note: Auto-validation is available only for specialized extraction models.
• Generate Data Type - If set to True, indicates that the output should be generated based on the selected extractor, resulting in an IDocumentData<ExtractorType> object. Alternatively, if set to False, indicates that the data generation should be skipped, resulting in a generic IDocumentData<DictionaryData> object.

  Visit Document Data for additional details and limitations available for the two object types.

Output

• Document Data - All the extracted field data from the file. Information can also be received from Classify Document.

  Visit Document data to learn how Document Data works and how to consume the extracted results for single and multi-value fields.

• App ID: Enter the App ID of the external application you previously created.
• App secret: Enter the App secret of the external application you previously created.
• Tenant URL: Enter the URL of the tenant where you created the external application. This is the tenant from where you will use resources at design-time.
  The URL should be in the following format: https://<baseURL>/<OrganizationName>/<TenantName>.

• Runtime Credentials Asset - Use this field when you need to access Document Understanding resources while the robot is connected to a local Orchestrator, or from a different tenant. You can choose to enter a Credential Asset, for authentication purposes, in one of the following ways:
  • From the dropdown list, select the desired Credential Asset from the Orchestrator to which the UiPath® Robot is connected to.
  • Manually enter the path to the Orchestrator Credential Asset where you store the external application credentials for accessing the project.
    The format of the path should be: <OrchestratorFolderName>/<AssetName>.
• Runtime Tenant Url - Use this field, alongside the Runtime Credentials Asset field. Enter the URL of the tenant that the robot will connect to in order to execute the extraction. The URL should be in the following format: https://<baseURL>/<OrganizationName>/<TenantName>.