Document Intelligence enables any organization to automate and accelerate the process of extracting data from documents. That data can easily be integrated into larger automation workflows to save time and resources.

Document Intelligence Xanadu

(What's new video: To come)

Data Normalization enhancement:

• Standardize extracted data by determining how to interpret ambiguous field values and by matching extracted values with data in a referenced table.
• Create a Reference field type to match extracted field values with the fields in a referenced table record.

Dashboards in the Admin experience: Monitor your Document Intelligence implementation with improved dashboard charts.

Document Classification integration workflow: Integrate document classification use cases with a custom application or workflow.

Additional Language support: Process documents that are written in Simplified Chinese or Japanese.

Document Intelligence Washington DC Updates

Version v4.0 and v4.1

Draw tool

Extract table information faster by selecting data directly from documents. Save time and reduce errors when extracting data from complex documents. The draw tool is available during extraction in the workspace.

Page range filter

Apply a filter to limit the pages that are sent for extraction. Send a better input to the model when not all the pages are required and address cases where multiple documents are combined in the same file.

Set up the page range filter as a parameter in the Attachment Filter of the Document Task.

Blank page

Classify blank pages in documents in the Document classification use cases for better classification outcomes.

Other enhancements

Increased limits - The file size limit for attachments is increased to 10 MB. The default maximum number of document tasks processed per instance per day is increased to 2,000.

General accessibility enhancements.

Improvements to AI quality for invoices - Improve extraction outcomes by leveraging more powerful pre-trained AI models, for invoices, with APO.

Document Intelligence Vancouver Updates

Version v3.1 (This version requires Vancouver) and Document Intelligence Admin v2.1

Guided field creation

Create new fields to extract from your documents easily in a new guided experience. Easily select the type of field from a single field, checkbox list, table, or group field. Easily create and manage multiple fields for group and table fields.

Date normalization

Create date fields to extract from your documents easily. Date fields are automatically converted to a date format and can be reviewed during extraction. The extracted value can seamlessly be stored in a date field on the target table.

Language support

Extract values from documents in Czech, Danish, Finnish, Norwegian and Swedish.

Version v3.0 (This version requires Vancouver) and Document Intelligence Admin v2.0

Document Classifier

Use AI to efficiently classify incoming documents and route them to the appropriate processing workflow. Learn more in the docs.

More flexibility around validation

It is possible to define which fields are required. Required fields must be filled during manual validation, or a warning message will be displayed. Only the required fields are needed to automate the extraction without agent validation. Fields that are not required may be left empty or unreviewed.

Note that straight-through processing is now called “fully automated mode”.

Improved model management

Make it easier to duplicate and export use cases, that includes the configuration and trained models.

Enhanced table extraction experience

Extracting values from tables is easier with new actions available: change the column size, multi-select, add a row above or below, clear values and delete rows.

The new table's schema requires a migration, follow Upgrade to Document Intelligence 3.0 in the docs.

Document Intelligence Utah Updates

Introducing the new Document Intelligence Admin, available from the store:

• A better user experience to set up, configure and monitor document processing solutions​

Learn more on this: Get Started with Document Intelligence Admin!

Version v2.3 and v2.4

• A new user experience for extracting information from tables (including better AI algorithms)
• Revamped terminology across the product
• Ability to process multiple documents concurrently
• Extract information from checkboxes for some use cases
• Extended the supported languages: English, French, German, Dutch, Spanish, Portuguese and Italian

Document Intelligence Tokyo Updates

Document Intelligence version v2.1 and v2.2

• Covers more use cases with support for table extraction
• Provides better AI for better information extraction outcomes
• Is easier to set up, for a faster time to value
• Offers a path to full automation with limited agent validation

Learn more

Document Intelligence Quick Start Guide

Introduction to Document Intelligence Data Extraction and Best Practices

Using Document Intelligence with Flow Designer

Document Intelligence FAQ

References

	Document Intelligence release notes