Release Notes

In this, we have introduced a validation summary, improved detection of the Final K-1 checkbox, enhancements to remaining functionality, dynamic container mapping with Encompass, customizable income PDF exports, optimized signal validation, inclusion of risk scores in SMB Analytics Excel exports, and improved deduplication logic.

What's new?

Capture

Validation summary

We’ve introduced a new validation summary feature in the Capture workflow to enhance data accuracy and improve automation readiness. The update includes support for both field-level and document-level validations and allows comprehensive checks across individual fields and multi-field relationships.

These rule-based validations are applied to the extracted data in both Complete and Instant workflows. These validations check field values against predefined formats, data ranges, and logical relationships to flag potential errors. Each validation includes a unique identifier (UUID), a descriptive message, and a passes_validation flag indicating whether the field passed the validation.

Validation results are returned directly through supported API endpoints to help you programmatically assess the quality of extracted data. This feature helps reduce manual review efforts and builds greater trust in automation, especially in Instant workflows where human verification is not involved.

👍

Info

The feature is currently available via the Ocrolus API and will be supported in the Dashboard in a future update.

To learn more, see the Validation summary. For enablement support, please get in touch with the Ocrolus support team.

Enhancements

Mortgage

Dynamic Container Mapping with Encompass (Beta)

The Dynamic Container Mapping feature enhances how mortgage documents are organized and indexed in the Encompass eFolder. This feature allows you to automatically group and route documents based on context-specific information, such as borrower name, financial institution, or property address, providing more precise control over document placement.

• Context-aware document grouping: Automatically groups and routes documents into specific eFolder containers using unique values (UVs) such as borrower name, financial institution, or property address.
• Improved borrower-level separation: Automatically organizes income and asset documents into individual folders for each borrower or co-borrower.
• REO versus subject property mapping: Separates documents related to real estate-owned (REO) properties from subject property documents using address-matching logic.
• Smart tax form grouping: Groups tax forms, such as Form 1040 and associated schedules, under a single parent folder based on form type and borrower.

This enhanced grouping applies automatically based on supported document types and lender configurations. However, lenders can still manually map or create documents.

Examples

• Asset documents, such as Assets: Bank Statement – Jill Smith – Chase and Assets: Bank Statement – Jill Smith – Bank of America, now appear in distinct folders.
• Income documents (W-2s, pay stubs, etc.) for each borrower are automatically grouped into clearly labeled folders, such as Income Documents – Jill Smith.
• Mortgage statements related to REO properties are routed separately from those related to the subject property.

🚧

This feature is in beta!

This feature is in beta. You may occasionally experience unannounced changes or bugs. We'd greatly appreciate your feedback on this feature and its accompanying documentation.

Income PDF export customization

The income PDF download now includes enhanced customization options to provide you greater control over the contents of exported PDFs. The key enhancements include:

• Custom selection: Select whether to include or exclude comments and source documents before exporting.
• Comments section: Optionally add a dedicated comments section to the exported PDF.
• Source documents: Choose to include or exclude supporting documents used for income calculations.
• Selective income export: Include only the selected income items; inactive employers are excluded by default.
• Smart defaults: Your previous selections are remembered using local storage, enabling quicker repeat exports.

Classify for Mortgage lenders - enhanced tax form grouping logic

We’ve upgraded our document processing capabilities to streamline your review experience and reduce manual effort. This enhancement offers the following benefits:

• Less manual cleanup, so you can trust the organization of your data and focus on higher-value tasks.
• Faster tax document reviews with fewer clicks and less sorting.
• Consistent document structure across all files for a better user experience.
• Improved grouping accuracy reduces the risk of misfiled or disjointed forms.

This upgrade enhances the accuracy of tax document identification and organization. The key improvements include:

• Year- and form type–based grouping, making it faster to locate specific pages.
• Automatic detection and grouping of key tax forms with their related pages, reducing manual sorting.
• Consolidation of state tax forms into a single PDF for easier access and review.

This enhancement is part of our ongoing commitment to simplifying document workflows and enabling faster decision-making.

Final K-1 checkbox detection on Schedule K-1 Forms (1065 and 1120S)

Ocrolus now identifies when the Final K-1 checkbox is selected on Schedule K-1 forms. Enabling this box indicates that the associated business is no longer active. In these scenarios:

• Income from that business will be excluded from the total monthly income calculation.
• An alert will display:

  Self-employment income was excluded due to the K-1 being marked as final.

This enhancement ensures more accurate assessments of borrower income.

Paystub multiplier - improved control over employment and leave periods

You now have greater control over employment and leave periods to ensure more accurate monthly income calculations:

• Pre-filled employment dates: Employment periods automatically populate based on WVOE or income documents.
• Easy editing: Quickly update employment dates or add multiple leave periods directly in the user interface.
• Automated calculation: Ocrolus automatically recalculates total months worked per year to adjust for any leave periods entered.

This update provides enhanced flexibility, resulting in more precise income calculations.

Reset functionality

We are streamlining the reset functionality on the Wage Earner page by removing the reset buttons from individual income types (Hourly, Salary, Overtime, etc.) and introducing two consolidated global options in the dropdown under the Save button. This improves clarity and consistency in how user edits are managed. Under the Save button, you now have two options:

• Reset Income: This will recalculate all income values back to the default calculations.
• Discard Unsaved Changes: This will discard any unsaved edits you have made available for all income types.

Detect

Optimized signal validation

The following five algorithmic signals have been fine-tuned to increase accuracy and reduce false positives related to pay stubs, W-2s, and dollar-amount edits in bank statements:

• Unreconciled gross pay – pay stubs: The following updates are made to reduce false positives for the Unreconciled gross pay signal:
  • Reduced sensitivity to minor rounding differences in pay stubs.
  • Adjusted the algorithm to better handle line items frequently excluded from total gross pay calculations, such as 401(k) matches.
• State taxes not paid – W-2s: The State taxes not paid signal is refined to accurately reflect states without state income tax. These states now include Alaska, Florida, Nevada, New Hampshire, South Dakota, Tennessee, Texas, Washington, and Wyoming.
• Suspected template – pay stubs: To enhance accuracy and reduce unnecessary alerts, the software used to detect paystub templates is recalibrated.
• Suspicious address – pay stubs: The address validation is improved to better distinguish minor formatting variations from genuinely suspicious addresses. Addresses not matching standardized U.S. postal records are flagged appropriately, with additional precision for non-geographic ZIP codes.
• Dollar amount edits – bank statements: The detection of Dollar amount edits for the specific bank statements is recalibrated to reduce false positive alerts and improve overall accuracy.

Analyze

Evaluate borrower default risk directly in Excel exports

The Ocrolus Risk Score can now be included in the SMB Analytics Excel export. This score provides a data-driven assessment of a borrower's likelihood to default, enabling underwriters to evaluate creditworthiness within the exported file quickly.

The Risk Score leverages over 500 Ocrolus cash flow features. It is trained on tens of thousands of matured business loans to detect patterns in transactional data that traditional credit models might overlook.

👍

Tip

You can also request custom risk scores tailored specifically to your organization's goals and borrower performance data.

Optimized de-duplication logic

The transaction deduplication logic is now optimized for more accurate matching based on transaction descriptions. Ocrolus can effectively identify and remove duplicate transactions, even when descriptions vary across multiple uploads into the same Book.

Classify and Capture

Additional attribute support

Ocrolus now offers support for the following attribute within the Marine Corps Basic Order document type configuration:

• iso_app-Part2-BusinessInfo:sicCode

To learn more, see the Marine Corps Basic Order page.