Latest Updates: June 2025

This article highlights the development work released to production throughout June 2025, including new features and enhancements, and bug fixes, made available to all users.

No Manual Upgrades Required

All updates to the Board Foresight application are version-controlled and automatically deployed, so customers always have access to the latest version without needing to manually upgrade.

Feature Releases and Enhancements

Single-Sign On

• Board Single Sign-On (SSO) is now enabled for joint Board/Foresight users.

Transformations

• Improved Transformation Order: Transformations now follow a strict sequence to ensure accuracy. Any frequency changes (e.g., converting monthly data to quarterly) must be applied before calculation-based transformations (e.g., Year-over-Year). For example, converting a monthly indicator to quarterly and then applying a YoY calculation is valid; the reverse order is not.

• Scenario Planning Enhancements: Users can now create and edit scenarios using Univariate transformed secondary indicators from a Model. Full forecast timelines are editable within the scenario.

• Transformation Carry-Through:

  • When launching a Discover job from the Indicator page, the transformed version of the indicator will now be passed through, ensuring consistency with user-applied settings.

  • Indicators with transformations applied on the Indicator page will now retain those transformations when added to a Workbench or a Model.

  • When a Custom Chart is created from a Workbench or Model, any applied transformations to indicators will automatically carry through to the chart.

• Centralized Charting Functionality: Chart options have been removed from the indicator page, as this functionality is now fully supported within the Transformations workflow.

Data and Discover Improvements

• CFTC Data Gap Handling: Weekly data sourced from the Commodity Futures Trading Commission (CFTC) will now have short data gaps (up to 2 weeks) automatically filled using the last available value. Gaps longer than 2 weeks will remain unfilled to preserve data integrity.

• Scoped Data Updates for Foresight Support: When Foresight Support assists with data integration, only the datasets included in the integration request will be updated. Manual uploads or changes made by other users within the same Company Instance will remain untouched unless explicitly included in the request.

• Improved Model Deletion Workflow: When locking a Model, the default prompt to create an indicator will now be toggled off. This change resolves a prior issue where users needed support team intervention to delete Models that had associated indicators created by default.

Insight Generation

• Exportable Model Insights: Forecast Fit and Forecast Error metrics are now available for export to Excel directly from the Model page, enabling easier analysis and sharing.

Bug Fixes

This update includes bug fixes related to: 

• Charting and Display

  • Indicators with one actual value and multiple forecast points now render correctly on charts.

  • Corrected a misspelling of "Econometric" on the Insights tab.

  • Updated the symbol and tooltip in the Workbench to accurately reflect sorting by R².

• Data Handling and Aggregation

  • Economic Forecast indicators managed by the Foresight Center of Excellence now aggregate correctly.

  • Index-specific data sourced from the Organization for Economic Co-operation and Development (also known as OECD) use Average instead of Sum for aggregation.

  • Seasonal profiles are now supported for aggregation.

• Workbench and Model Experience

  • When an indicator is transformed, the statistical relationships (e.g., correlation, coefficients) automatically update on both the Workbench and Model pages.

  • If a shortened seasonality test window was previously applied, users returning to adjust the date will now see the full data history allowance, not the restricted view from the earlier run.

  • A user-friendly message is now displayed when Relative Importance fails to load due to the two most common causes. This allows users to identify and resolve the issue directly.

• GenAI: GenAI commentary in Model output now respects the correct lead time defined by the model.