Support for additional Planning Analytics data source types

Additional data source types are now supported for Planning Analytics data source connections on IBM Planning Analytics as a Service and Planning Analytics Advanced Certified Containers. 

You can now connect to the following data source types: Amazon Redshift, Box, Dremio, Dropbox, FTP, Google Cloud Storage, IBM Netezza Performance Server, and Microsoft Azure Fabric Warehouse.

Amazon Bedrock model provider support in Planning Analytics Administration

Administrators can now configure Amazon Bedrock as the AI model provider for AI-generated insights directly from Planning Analytics Administration.

The Amazon Bedrock model provider option is now available in the AI configuration settings in Planning Analytics Administration.

This enhancement allows administrators to switch between watsonx.ai and Amazon Bedrock without manually editing configuration files.

• The default AI model provider is watsonx.ai.
• When Amazon Bedrock is selected as the model provider, administrators can configure the corresponding Bedrock credentials and model settings.

The Amazon Bedrock model provider option is controlled by the showBedrockModel feature flag.

Use Planning Analytics Workspace REST API for database management

You can manage TM1 databases programmatically by using new REST API endpoints in Planning Analytics Workspace.
New REST API endpoints are available to manage TM1 databases without using the user interface. These endpoints support the following operations:

• List all TM1 databases
• Start, stop, and restart TM1 databases
• Create and delete TM1 12 databases
• Manage and access TM1 12 database backups

Note: Create, delete, and backup management operations are available for TM1 12 databases only.

Enhancements to the Exploration analysis report

The Exploration analysis report now includes seven structured sections, richer visualizations, and Learn moredocumentation links to help you interpret results more quickly.
The report is organized into the following sections:

• Data Overview — Descriptive statistics table and mean-value bar chart.
• Impact Analysis — Redesigned bubble chart with distinct colors and non-overlapping labels. A context note identifies the top N attributes that drive subsequent sections.
• Trend & Volatility — Horizontal bar chart showing trend strength and volatility side by side for each feature.
• Feature Correlation Heatmap — Pearson correlation heatmap across the top key features.
• Relative Impact Analysis — Dedicated page per feature with improved descriptions and a fixed network graph.
• Series Decomposition (STL) — Trend, seasonal, and residual components with cycle detection per feature.
• Anomaly Detection — Scatter and line charts that highlight anomalies per feature. A note is shown when no anomalies are detected.

Other improvements include chart captions below every chart and sharper PDF output at a standardized resolution.

Extended forecasting engine

You can now use the Extended forecasting engine to access advanced algorithms that support more forecast use cases, such as multiple seasonality patterns or high-frequency daily forecasting.

In the forecasting Advanced settings, administrators and modelers can select Standard or Extended under Algorithm engine. If you select Standard, the current forecast engine is used. If you select Extended, you can select from advanced algorithms for univariate and multivariate forecasting. Algorithms and their parameters can be selected automatically or manually. The engine selection is available for both on-demand forecasting and baseline forecast assets.