Overview

The Reconcile Doc Extraction node allows you to assign an interface to review and update extracted data from a doc extraction run, started by the started by the Extract from Document smart service, which leverages the Document Extraction AI Skill.

The Extract from Document smart service offers a simplified development experience for extracting data from a document. The Reconcile Doc Extraction smart service can also be used with additional document extraction functionality in a process model. See the following list for the other smart services and functions you can use with the Reconcile Doc Extraction Smart Service:

• Start Doc Extraction Smart Service
• a!docExtractionStatus()
• a!docExtractionResult()
• Admin Console > Document Extraction

Properties

• Category: Document Management

• Icon:

• Assignment Options: Attended/Unattended

Configuration options

This section contains tab configuration details specific to this smart service. For more information about common configurations see the Process Node Properties page.

Data tab

This tab displays all the Node Inputs and Node Outputs for the Service Node. You can add additional inputs and outputs, if needed.

Node inputs

The default inputs are:

Input	Data Type	Required	Multiple	Description
Document Extraction Id	Any Type	Yes	No	ID of the doc extraction run. Usage with Extract from Document smart service: use the Doc Extraction Id output from the Extract from Document smart service in this field. Usage with Start Doc Extraction smart service: the a!docExtractionStatus() function should return COMPLETE. Then, use the Doc Extraction Id output from the Start Doc Extraction smart service in this field.
Data Type Number	Integer	Usage with Extract from Document smart service: No. Usage with Start Doc Extraction smart service: Yes.	No	Usage with Extract from Document smart service: leave this value empty. The document structure is defined via the AI Skill you choose in the Extract from Document smart service. Usage with Start Doc Extraction smart service: Data Type Number is the type to which the values should be cast. This should be a Custom Data Type (CDT). This may be retrieved by using typeof(x), where x is a value of the intended type.
Title	Text	No	No	Text for the title of the interface. Default is Reconcile [name of document].
Submit Button Label	Text	No	No	Label for the "Submit" button. Default is Submit.
Mark As Invalid Button Label	Text	No	No	Label for the "Mark as Invalid" button. Default is Mark as Invalid.
Is Structured Doc	Boolean	No	No	Set to true when fields to be extracted are in the same location for all documents. When you pass in structured documents and perform reconciliation, Appian learns to recognize this information based on its position, improving results for future extractions.

Node outputs

Name	Data Type	Description
Reconciled Data	Any Type	The data reconciled by the user. Usage with Extract from Document smart service: returns the extracted data in the form of the data structure defined in the AI Skill. Usage with Start Doc Extraction smart service: the data type will match the data type number passed in the inputs.
Is Submit	Boolean	User selected the "Submit" button.
Is Invalid	Boolean	User selected the "Mark as Invalid" button.

To save the Reconciled Data into a process variable:

1. Select the Outputs property sheet (tab).
2. Click Reconciled Data on the Results list. The result properties are displayed.
3. Create a new process variable to store the result, or select an existing process variable with data type of your input Data Type Number from the available list.

Usage considerations

The doc extraction reconciliation interface can only be accessed by a user who has permissions to the original document.

Appian will delete document extraction runs after 30 days, or when the total amount of data surpasses 10 GB. If you attempt to access a run that has been deleted, you will see an error. Appian will not delete the documents you uploaded.