What's new in IBM Software Hub

Starting in Version 5.4, backup and restore operations run on the Red Hat® OpenShift® Container Platform cluster for improved reliability and scalability.

In addition, the cluster administrator can give instance administrators the ability to manage backup and restore operations for an instance of IBM Software Hub, which can help you streamline operations and improve efficiency.

Tech preview This is a technology preview and is not supported for use in production environments.

View and manage data source definitions in connection forms

You can now view and refresh Data Source Definitions (DSDs) directly within the connection form. The interface shows the status of DSD lookup and integrates with Test connection, making it easier for you to verify and manage connection details during setup.

Parametrize connection properties

You can now parametrize connection properties, including boolean values, by linking connections to parameter sets or binding individual properties. This lets you manage configuration centrally and reuse connections across environments without modifying them.

Control access to data in spaces

You can now control whether users or groups can view or preview actual data in spaces by using platform permissions. This helps you restrict access to sensitive data across assets, including previews and data‑related tools.

Run and manage jobs with the Job Operator role

You can now use the Job Operator role in Deployment Spaces to run, schedule, and monitor jobs without modifying pipelines or other assets. This allows you to perform operational tasks while restricting access to design‑time changes, improving control in production environments.

Specify model modalities for custom foundation models

You can now define model tasks (modalities) such as text, image, audio, or video when importing custom foundation models. This helps you clearly indicate supported use cases and adjust modalities at deployment time to match your requirements.

Version 13.0.0 of the common core services includes various fixes.

For details, see What's new and changed in the common core services.

Configure guaranteed resources for a project

If the scheduling service is installed, you can configure guaranteed resources for a project from the web client. When you enable guaranteed resources for a project, the scheduling service reserves the allocated resources to run assets associated with this project. You might need to wait for the resources to become available if the

If you configure guaranteed resources, keep the following restrictions in mind:

• The resources might not be available immediately. You might need to wait until there are sufficient resources to fulfill the request.
• The request might prevent other processes from starting to ensure that the resources are available to the project.

Version 2.0.0 of the scheduling service includes various fixes.

For details, see What's new and changed in the scheduling service.

Version 5.4.0 of the AI Factsheets service includes various fixes.

For details, see What's new and changed in AI Factsheets.

For details, see What's new and changed in Analytics Engine powered by Apache Spark.

Updated software version for Cognos Analytics

This release of the service provides Version 12.1.2 of the Cognos Analytics software. For details, see Release 12.1.2 in the Cognos Analytics documentation.

Version 30.0.0 of the Cognos Analytics service includes various fixes.

For details, see What's new and changed in Cognos Analytics.

Data Gate supports TLS 1.3

Data Gate now supports TLS 1.3 for connections between Db2 for z/OS and Data Gate, which eliminates less safe legacy cryptographic algorithms and reduces handshake overhead to a single round trip. Every session benefits from forward secrecy by default, which satisfies compliance requirements that mandate newer TLS versions. Customers can choose between TLS 1.2 and TLS 1.3 by configuring their AT-TLS rules accordingly. See Configuring network access between Data Gate and IBM Z.

New REST APIs for programmatic management of replicated tables

A new set of REST APIs provides an initial set of operations for retrieving information about and working with replicated tables, enabling automation through CI/CD pipelines, infrastructure-as-code, or custom tooling without touching the UI. See Managing Data Gate tables.

Monitoring via a Prometheus-compatible metrics endpoint

Data Gate exposes a Prometheus scrape endpoint, making key metrics — synchronization latency and throughput — available to any compatible monitoring stack. Teams using Grafana can plug straight in for dashboards and alerting. This feature replaces the deprecated monitoring UI. See Monitoring a Data Gate instance.

Configurable database name for remote Db2 target databases

Previously, Data Gate enforced a 1:1 relationship between a Data Gate instance and its target database, with the target database name hardcoded as BLUDB across multiple components. Starting in this release, the target database name is configurable for remote Db2 and Db2 Warehouse scenarios. You can now specify the target database name when it differs from BLUDB. This also enables the possibility of reusing an existing Db2 instance — including instances with multiple databases — as the target for more than one Data Gate instance. See Connecting to a remote Db2 instance

Query acceleration and storage saver for remote Db2 target databases

Data Gate can be connected to a remote Db2 (transactional) or remote Db2 Warehouse (analytic) instance. For analytic targets, Data Gate supports a special mode that enables additional use cases from the Db2 for z/OS source side. Starting in this release, query acceleration mode is now supported for remote Db2 targets. Query acceleration mode allows you to execute queries via Db2 for z/OS on Data Gate tables, and to archive table partitions to move them to Data Gate (storage saver). This option can be selected at instance provisioning time. Extending support to remote Db2 targets brings feature parity with other target configurations. See these links for more information: Data Gate query acceleration and Data Gate storage saver.

Helm-based management of Data Gate instances without Operator Lifecycle Manager

Data Gate instances can now be managed via Helm charts, which unlocks GitOps-based instance management workflows and provides a better tracking of deployment phases via improved progress messages.

Version 15.4.0 of the Data Gate service includes various fixes.

For details, see What's new and changed in Data Gate.

Version 5.4.0 of the Data Privacy service includes various fixes.

For details, see What's new and changed in Data Privacy.

Access data assets with AWS Databricks delivery method

You can now use the AWS Databricks delivery method to access a data product with AWS Databricks assets. After you subscribe to a data product and it is delivered to you, you can directly access the AWS Databricks resources and view details about how to use them.

Access data assets with Snowflake delivery method

You can now use the new Snowflake delivery method to access a data product with Snowflake assets. After you subscribe to a data product and it is delivered to you, you can directly access the Snowflake resources and view details about how to use them.

Integrate notebooks across watsonx.data intelligence and projects

When you install Watson Studio or watsonx.ai, you can use Data Product Hub to share or reuse your watsonx.data intelligence notebooks. As a data producer, you can add notebooks and data assets from your project directly to a data product. When you deliver a data product, recipients can download the included notebooks for their own use. You can also add notebooks to projects from data products.

Customize data product request forms to match your organization's needs

Customize your data product request forms to collect the exact information that your organization requires. Choose and edit multiple field types to create a form that fits your workflow.

Create flexible data product delivery options with custom delivery methods

Create a custom delivery method based on your use case. You can configure your delivery method to be used in restricted or public data products and to support multiple data source types and asset types.

Version 5.4.0 of the Data Product Hub service includes various fixes.

For details, see What's new and changed in Data Product Hub.

Use new attribute to create Data Refinery flows externally without the UI

You can now use the new shaperAPICreated attribute to create Data Refinery flows programmatically without needing to use the UI. This capability means that you can:

• Use external APIs to create Data Refinery flows.
• Use third-party integrations to generate flows with shaping operations.
• Use automated workflows to create data transformation pipelines.
• Use custom applications to build Data Refinery flows without using the UI.

For more information, see the API documentation: <hostname:port_number>/v2/data_flow_spark/docs/swagger/index.html

Create Data Refinery flows in folders

You can now create Data Refinery flows in folders or save existing flows in folders. The information panel shows the folder paths for the flow and the target that you chose. You can also create jobs in folders and modify the flow and target folder paths in the flow settings.

Define parameters for source and target data in Data Refinery flows

A new parameter step is now available in the job creation wizard for Data Refinery flows. You can define parameters for both source and target data so that the same job can be used with different data sets. You can also edit existing jobs to use parameters to define source and target data.

Cancel Data Refinery jobs in "starting" state

You can now cancel Data Refinery jobs that are in the Starting state. This enhancement improves job management and resource control.

New connections for Data Refinery

You can now use the following connections with Data Refinery:

• Vertica
• Microsoft Azure Databricks

Version 13.0.0 of the Data Refinery service includes various fixes.

For details, see What's new and changed in Data Refinery.

Version 5.4.0 of the Data Replication service includes various fixes.

For details, see What's new and changed in Data Replication.

Connect to AlloyDB for PostgreSQL databases

You can now use the AlloyDB for PostgreSQL connector in your DataStage flows to read and write data from AlloyDB for PostgreSQL databases.

Access data in AWS Databricks

You can now use the AWS Databricks connector in your DataStage flows to access and process data in Databricks workspaces.

Access files in Microsoft SharePoint

You can now use the Microsoft SharePoint Files on Canvas connector in your DataStage flows to read and write files stored in SharePoint document libraries.

Access data in Microsoft Dynamics 365

You can now use the Microsoft Dynamics 365 connector in your DataStage flows to read and write business data from Dynamics 365 applications.

Export and import compiled pipeline binaries

You can now export and import compiled Python binaries with optimized runner pipelines, which means that you can move pipelines together with their compiled assets. You control this behavior by using the include-python-binaries and include-common-binaries options in cpdctl.

Data encryption for Teradata connections

You can now enable full session data encryption for Teradata optimized flows by using the new Data Encryption option. This option uses either TDGSS or TLS/SSL to encrypt network traffic, SQL statements, data requests, and responses for the entire session.

Create parameter sets from connection properties

You can now create parameter sets directly from connection properties for supported connectors. Select one or more connection types and add their properties as parameters so that you can easily reuse and manage configuration values across pipelines.

Run remote engines on s390x remote engines

You can now run remote engines on s390x (IBM Z and LinuxONE) systems, deployed as Docker containers or in Kubernetes clusters. This allows you to submit jobs from x86_64 environments and execute them on s390x hardware. This capability enables workload distribution across heterogeneous architectures.

Version 5.4.0 of the DataStage service includes various fixes.

For details, see What's new and changed in DataStage.

Data Virtualization attempts to publish objects to catalogs unless explicitly skipped

When you publish objects from Data Virtualization to a catalog, Data Virtualization now always attempts to publish any duplicate asset to the target catalog. You can override this behavior by either selecting Skip publish if assets already exist in the catalog or by passing "allow_duplicates": false in the REST API payload. If duplicates already exist and you do not select to skip, then what occurs next is determined by the Duplicate asset handling setting on the catalog. Catalogs are configured with Update original assets by default, where existing copies are updated while preserving governance annotations like business terms and data classes. To change the duplicate asset setting, see Duplicate asset handling methods.

Apply consistent asset names across Data Virtualization assets in catalogs and projects

When you publish objects from Data Virtualization to catalogs and projects, the resulting asset name no longer includes a schema prefix, which ensures consistent naming across Cloud Pak for Data. This change minimizes the possibility of duplicate assets when you import, enrich, and update assets that are published from Data Virtualization. In addition, you can also remove schema prefixes from your Data Virtualization asset names across all your catalogs, projects, and spaces by running the REMOVESCHEMAFROMASSETS stored procedure. When you run this procedure, Data Virtualization scans all the existing assets and renames them without the specified schema prefix.

To rename your assets to remove schema prefixes, see the REMOVESCHEMAFROMASSETS stored procedure from Data Virtualization procedures.

Control how connections are mapped during import operations with connection ID (CID) mapping management 

Before you import your Data Virtualization objects, you can now choose to create new CID mappings or update how existing connections are mapped from your source environment to your target environment by running the MANAGE_CID_MAPPING stored procedure. You can run this procedure as part of both the Git based and Data Virtualization API based migration workflows.

See Optional: Managing connection mappings during import.

Use OAuth 2.0 authentication when you create a REST API connection

You can now use the Open Authorization (OAuth) 2.0 authorization protocol for authentication when you create a new REST API connection from the Data Virtualization web client. This option is now available in addition to existing Basic Authentication and Bearer Token options.

Deny access to virtual tables while masking the derived views with IBM Knowledge Catalog data protection rules

You can now use IBM Knowledge Catalog data protection rules to deny a user access to virtual tables while masking the derived views. You can apply deny and masking rules simultaneously and follow the Db2 authorizations plus RCAC model applicable to views.

Deny access to objects that are not published to a governed catalog by enabling the Restrict access to ungoverned objects setting

Data Virtualization Administrators can now enforce governance on all objects by configuring Data Virtualization to deny access to any objects that are not published to a governed catalog.

Caches now automatically clear when you change your personal credentials setting

You can now change your credential settings from personal to shared credentials, or vice versa, even after creating virtual tables or caches, without manually removing any caches linked to the connection.

To change your credentials, see the TOGGLEPERSONALCREDENTIALSUSEINCACHE and USE_PERSONAL_CREDENTIALS stored procedures in Data Virtualization procedures.

Use Instana® for better data observability

You can now send your metrics to Instana to monitor and analyze Data Virtualization. To enable Instana on your Data Virtualization service or on each of your Data Virtualization instances.

See Enabling Instana for Software Hub.

Version 3.4.0 of the Data Virtualization service includes various fixes.

For details, see What's new and changed in Data Virtualization.

Accelerate AI workloads with IBM Db2 12.1.5

You can now run AI-driven workloads more efficiently with the Db2 service on IBM Software Hub. This service is built on IBM Db2 12.1.5, which includes advanced vector processing and index-based acceleration to speed up similarity searches in both traditional and containerized environments.

Version 5.4.0 of the Db2 service includes various fixes.

For details, see What's new and changed in Db2.

Version 8.5.0 of the Db2 Big SQL service includes various fixes.

For details, see What's new and changed in Db2 Big SQL.

Version 5.4.0 of the Db2 Data Management Console service includes various fixes.

For details, see What's new and changed in Db2 Data Management Console.

Accelerate AI workloads with IBM Db2 12.1.5

You can now run AI-driven workloads more efficiently with the Db2 Warehouse service on IBM Software Hub. This service is built on IBM Db2 12.1.5, which includes advanced vector processing and index-based acceleration to speed up similarity searches in both traditional and containerized environments.

Version 5.4.0 of the Db2 Warehouse service includes various fixes.

For details, see What's new and changed in Db2 Warehouse.

Compare and visualize scenario differences in a Decision Optimization experiment

You can now compare and visualize differences and similarities between two scenarios in a Decision Optimization experiment. By comparing scenarios, you can better understand how different model formulations, data, and parameters impact your optimization results.

Version 13.0.0 of the Decision Optimization service includes various fixes and updates.

For details, see What's new and changed in Decision Optimization.

This release of the EDB Postgres service includes various fixes.

For details, see What's new and changed in EDB Postgres.

Version 5.4.0 of the Execution Engine for Apache Hadoop service includes various fixes.

For details, see What's new and changed in Execution Engine for Apache Hadoop.

Run multiple data quality rules with a single click

To run multiple data quality rules at once, you can now organize them in data quality rule groups.

Import, enrich, and assess the quality of data from additional data sources

You can now import metadata from Microsoft Azure Fabric Warehouse data sources, enrich that data, and assess its quality.

Query data assets and document libraries in natural languages other than English

The Text-to-SQL service now uses a multilingual embedding model so that you can create natural language queries in languages other than English. In upgraded deployments, you must reprocess existing metadata with the new model.

Sync reference data sets to external databases

Reference data sets can now be synchronized to external databases and then consumed as regular governed data assets. With this enhancement, a reference data set can be profiled, queried via SQL, joined with other data sets, and reused consistently across the platform. This enables broader consumption of reference data for validation, standardization, analytics, and AI use cases.

Import and export reporting queries

You can now export your reporting queries in batch into a JSON file, and import JSON files with queries. Bulk import and export lets teams easily migrate, back up, and manage large collections of reporting queries across environments.

Main menu improvements

You can now view and manage all artifacts grouped by type, and all categories by opening Governance > All artifacts from the main menu.

Control data and metadata access with the Access data preview permission

You can now assign the Access data preview permission to users and user groups across all projects, catalogs, and deployment spaces to have more control over who can view the actual data and address security requirements. Users without this permission can view asset metadata, but they are blocked from previewing actual data.

By default, the new permission is included in the following predefined roles:

• Administrator
• Business Analyst
• Data Engineer
• Data Scientist
• Data Steward
• Data Quality Analyst
• Developer
• User

If watsonx.ai and DataStage are deployed, the following tools and views are disabled for users and user groups without the permission:

Watsonx.ai

• AutoAI
• SPSS Modeler
• Jupyter
• Pipeline
• RStudio

DataStage

• Data previews (canvas)
• Data previews for data sets and file sets

Update role assignments and any custom roles that you have for users who need to preview asset data.

Browse asset hierarchies

You can now explore and manage connected, custom, and unstructured assets from the Hierarchies pane on the catalog Assets page. Information about relationships between available data sources, databases, schemas, tables, and columns in catalogs is available in one structured view. With these details, you can, for example, navigate relationships across multiple asset levels, view detailed information for assets at any level, or bulk select assets across levels to manage metadata more efficiently.

Version 5.4.0 of the IBM Knowledge Catalog service includes various fixes.

For details, see What's new and changed in IBM Knowledge Catalog.

Track data flow changes by comparing lineage versions

You can now compare two versions of a lineage graph to identify which assets were added, removed, or modified between versions. This comparison helps you to understand how your data flows evolved and assess the downstream impact of those changes.

Monitor processed OpenLineage events

You can now monitor OpenLineage events in a centralized dashboard to verify event ingestion, identify failed or pending events, and troubleshoot processing issues. The dashboard also helps you understand the overall health of the OpenLineage processing pipeline by showing event volume and trends over time.

Export data lineage to OpenLineage payloads

You can now export data lineage as an OpenLineage .json payload, making it compatible with any application that supports the OpenLineage standard format.

Import lineage from new data sources

You can now import lineage metadata from the following additional data sources:

• IBM Netezza Performance Server
• Tech preview SAP BusinessObjects
• SAP HANA

Connect to new data sources by using version 1.5.0 of the Manta agent

You can now import lineage metadata from the following data sources by using the updated Manta agent:

• Db2
• Db2 for z/OS
• Db2 on Cloud
• SAP BusinessObjects
• Tableau

Manta agent version 1.5.0 is now available. Version 1.4.0 is deprecated, and versions 1.0.0, 1.1.0, and 1.2.0 are no longer supported. Update unsupported agent versions and consider upgrading version 1.4.0 instances to the latest version 1.5.0.

Pause and resume the import of lineage metadata from Qlik Sense and SAP HANA

You can now pause lineage metadata imports from Qlik Sense and SAP HANA and resume them at your convenience. This control is especially useful when importing large data volumes that require extended processing time.

Delete lineage from a specific point in time

You can now delete lineage data from a specific time range to remove data that you no longer need or that was imported by mistake. Only data imported during your specified time range is deleted, helping you focus on relevant lineage information.

View asset ownership directly on the lineage graph

You can now see the assigned owner, whether a user or a group, displayed directly on assets in the lineage graph. This immediate visibility saves you time by eliminating the need to open the metadata details panel to find ownership information as you explore data lineage relationships.

Version 5.4.0 of the IBM Manta Data Lineage service includes various fixes.

For details, see What's new and changed in IBM Manta Data Lineage.

Optimize matching algorithms by using enhanced pair analysis and recommendations

You can now create and manage multiple pair analysis tasks simultaneously and use their results to generate tuning recommendations for your matching algorithm. With these enhancements, you now have greater flexibility in your matching and algorithm tuning workflow. Additionally, you can now:

• Generate algorithm tuning recommendations before data stewards complete all pair reviews.
• Include manual stewardship decisions (such as link, unlink, and potential match decisions) in the calculations that generate your tuning recommendations.
• Visualize tuning outcomes by using confusion matrices, histograms, and pie charts to understand how changes will impact your algorithm before you apply them.
• Request new pair analysis tasks while others remain in progress.
• Delete unnecessary pair analysis tasks or results.

Automate relationship creation in your master data by defining discovery rules

You can now configure discovery rules to automatically establish and maintain relationships between your master data records and entities. Configure conditions and filters that evaluate your master data to discover and create relationships. Discovery rules work across record-to-record, record-to-entity, and entity-to-entity relationships, ensuring that your relationships stay current and consistent.

Find and fix MDM data quality issues more easily

You can now search for and resolve data quality issues in your master data by using the Stewardship tab. Review potential matches that need manual linking decisions and potential overlays that might indicate incorrect record updates. Click an issue to start working on the remediation task immediately. The new streamlined view helps you focus on the issues that need attention so that you can maintain more accurate master data.

Organize MDM data quality issues by using custom tags

You can now create and manage custom tags in IBM Master Data Management to help your team quickly find, identify, and prioritize data quality issues in your master data. Tags work as searchable, color-coded labels that make it easier to organize and track issues across your organization. You can also configure potential overlay workflows to automatically apply tags to the issues that they create, so your data stewardship processes are streamlined.

Protect sensitive data by configuring access to master data

You can now configure how IBM Master Data Management controls user access to data. Access control strategies ensure that only authorized users can access sensitive or confidential information, such as personally identifiable information (PII).

Configure one or both of the following access control types:

• Attribute-based access control (ABAC) protects specific data characteristics, across all data, from unauthorized users.
• Token-based access control (TBAC) uses security tokens to define user access at the row level for each record.

Modernize your master data management by migrating from InfoSphere® MDM Standard Edition or Advanced Edition or InfoSphere Big Match

You can now migrate your existing master data and matching algorithms from IBM InfoSphere Master Data Management (InfoSphere MDM) or IBM InfoSphere Big Match for Hadoop to the IBM Master Data Management service. By migrating, you gain access to modern, cloud-native capabilities and integration with other IBM Software Hub services.

The IBM Master Data Management migration service provides easy-to-use APIs that preserve your data structure and integrity while minimizing downtime. Your master data entities, relationships, groups, and matching algorithms remain intact throughout the migration process. During migration, both systems run in parallel so that you can validate that the service works correctly before you start using it as your production MDM solution.

Control master data entity attribute composition at the field level

When you configure attribute composition rules, you can now define filtering and prioritization logic at the field level by using value-based rules. As a result, you now have finer control over which record attribute values get surfaced to the entity.

Value-based composition rules help you filter low-quality data and construct more accurate, business-aligned entities. You can exclude invalid values like placeholders or dummy data, prioritize specific values in custom order, apply comparison functions to select for conditions such as the longest name or highest score, and create conditional rules that change depending on data conditions.

Prioritize rare matches over common ones in your master data

You can now configure your IBM Master Data Management matching algorithm to score matches based on how common or rare the matched values are in your actual dataset. The algorithm uses your real data distribution to boost scores for distinctive matches and reduce scores for common ones.

For example, matching on the rare last name "Xylander" should score higher than matching on the common name "Smith," because rare matches are more likely to identify the same person. This prevents the algorithm from over-scoring matches on common values like "John Smith" while under-scoring matches on distinctive values like "Hamish Xylander."

Validate and enrich reference data by using code tables

You can now centralize and manage all of your reference data (such as country codes, status values, and product categories) by using code tables. Code tables automatically validate data as it enters your system, store data efficiently by removing redundant display values, and enrich data with human-readable labels in your preferred language when you retrieve it. Code tables also support multiple languages with automatic fallback, ensure data consistency across your application, and let you update reference values without deploying code changes.

Exchange patient data with HL7-enabled healthcare systems

You can now exchange patient data between IBM Master Data Management and your HL7-enabled healthcare systems by using a message broker. Use the HL7 message broker to maintain a single source of truth for patient information across hospital registration systems, electronic medical records, laboratory systems, and other healthcare applications that use the HL7 communication protocol.

This capability is specific to the healthcare industry and is not enabled by default. For more information, see the IBM Master Data Management administration topics.

Version 4.12.43 of the IBM Master Data Management service includes various fixes.

For details, see What's new and changed in IBM Master Data Management.

Access Amazon MSK from a pipeline by using IAM access control

When you deploy a Data Collector 7.4.0 engine, you can connect to Amazon Managed Streaming for Apache Kafka (Amazon MSK) by configuring a custom authentication option in your pipeline.

For the complete list of new features in Data Collector 7.4.0, see 7.4.x Release Notes in the IBM StreamSets documentation.

Deploy an engine to the embedded data plane on the IBM Software Hub cluster

To help you get started with IBM StreamSets, you can now deploy a Data Collector engine to the same cluster where IBM Software Hub is installed.

Each organization includes a default environment that is named Embedded Data Plane that represents the Kubernetes cluster within Red Hat OpenShift where IBM Software Hub is installed. You can activate this environment and then create Control Hub Kubernetes deployments to automatically provision Data Collector engines that run in the cluster.

For more information, see Post-installation tasks in the IBM StreamSets documentation.

Version 6.4.0 of the IBM StreamSets service includes various fixes.

For details, see What's new and changed in IBM StreamSets.

Version 10.2.0 of the Informix service includes various fixes.

For details, see Fix list for Informix Server 15.0.0.2 release.

Version 42.16.0 of the MANTA Automated Data Lineage service includes various fixes.

For details, see What's new and changed in MANTA Automated Data Lineage.

Use watsonx.ai for translation services on OpenPages on IBM Software Hub

You can now use watsonx.ai as a translator type for translation services on OpenPages on IBM Software Hub.

Create and manage Network Policies with the optional Role-Based Access Control

You can now use optional Role-Based Access Control (RBAC) for Network Policies in non-OLM (Helm-based) deployments. This feature gives administrators control over whether the operator can create and manage Network Policy resources. By default, Network Policies are enabled (enableNetworkpolicies: true).

The operator's RBAC roles should only have read permissions for Network Policies, such as get, list, and watch.

Add favorite canvases to your dashboard

You can now mark canvases as favorites and display them in a new dashboard panel called Favorite Canvases. You can add favorite canvases in the user and admin dashboards for easier access and enhance productivity.

For more information, see Adding a Favorite Canvases panel.

Preview attachments in OpenPages

You can now view PDF, text, and image files directly in OpenPages without downloading them. Files are rendered in an OpenPages tab to ensure content is no longer accessible when the session expires.

The preview feature improves document accessibility and security by allowing users to preview attachments without leaving the OpenPages interface or downloading files to their local system.

For more information, see Adding and managing all files (attachments).

View object date fields in user's time zone

Date fields are now displayed in the user's time zone. Unlike timestamp fields, date fields follow the system date time zone support and ensures consistent date display across the application.

For more information on date fields, see Understanding system and non‑system date fields

For more information on configuring the setting, see Time zone conversion for non‑system date fields.

Access an embedded watsonx Orchestrate chat within OpenPages

You can now access an embedded watsonx Orchestrate chat within OpenPages. Without leaving the platform, you can quickly interact with AI agents, which can use tools from the OpenPages MCP server and other MCP servers.

Create more complex assessment flows with six levels of dependent questions

Questionnaire templates now accommodate up to six levels of dependent questions, increased from the previous three-level limit. You can create more complex assessment flows with deeper question hierarchies.

Schedule move and rename jobs

You can now schedule entity and non-entity move jobs for a custom date and time. When you schedule these jobs, you have more flexibility, especially when you move or rename large data sets or when you need to start a job instantly.

For more information, see Scheduling move and rename jobs.

Version 10.0.0 of the OpenPages service includes various fixes.

For details, see What's new and changed in OpenPages.

Manage and standardize configuration by using tenant-level settings

You can now configure Orchestration Pipeline settings at the tenant level, in addition to pipeline and project levels. This lets you manage and standardize configuration across pipelines using cpdctl, making it easier to enforce consistent behavior in multi‑tenant environments.

Simplify pipeline runs by using inline mode for standard runtime

You can now run pipelines in inline mode using the standard runtime. By using inline mode, you can avoid creating separate jobs for nested pipelines and DataStage runs, simplifying job execution and pipeline management.

Standardize pipeline job names across your tenant

You can now apply consistent job names to pipeline runs by using a single configuration setting. Set the job naming behavior once at the tenant level by using cpdctl to apply consistent naming across all pipelines without requiring updates in individual projects.

Analyze pipeline runs more easily by viewing separate logs for nested jobs

You can now review pipeline execution more easily by using separate logs for nested jobs. When you run pipelines, logs from subordinate jobs are now stored separately instead of being included in the pipeline runner log. This separation makes it easier for you to find, view, and analyze the log details for each job.

Grant operational pipeline access without edit permissions

You can now manage pipeline operations by using the new job operator role. Job operator users can run pipeline jobs and reset the job run cache without modifying pipeline configurations. This new role gives you a way to grant operational access without also granting permission to edit pipelines.

Run StreamSets job component

You can now run StreamSets jobs directly within pipelines by using the Run StreamSets job component. This allows you to integrate StreamSets job execution into pipeline workflows.

Version 5.4.0 of the Orchestration Pipelines service includes various fixes.

For details, see What's new and changed in Orchestration Pipelines.

Version 5.4.0 of the Planning Analytics service includes various fixes.

For details, see What's new and changed in Planning Analytics.

Access management and Organization explorer feature

Authorized administrators can now more easily manage and organize users, roles, access control groups, and organizations for application users.

Administrators can now easily:

• Manage application users: Create, view, update, enable, disable, archive, and reactivate users.
• Manage roles: Create, view, and update user roles.
• Manage access control groups: Create, view, and update access control groups and manage object-to-access-control-group mappings.
• Manage organizations: Create, view, edit, and delete organizations. For details, see. Using Organization explorer.

New capabilities for the Suspect duplicate processing feature

The process of duplicate matching and merging data based on the data within the catalog is now improved so that you can automatically:

• Bulk identify and mark duplicate records.
• Review and act on the duplicates for the multiple records.
• Run a catalog-wide duplicate match through a report job.

Natural Language Processing (NLP)-based search

Catalog users can now search data across all catalogs by using natural language queries.

Version 10.0.0 of the Product Master service includes various fixes.

For details, see What's new and changed in Product Master.

Version 13.0.0 of the RStudio Server Runtimes service includes various fixes.

For details, see What's new and changed in RStudio Server Runtimes.

Experiment with TLA rules in the Rule lab

You can now validate and refine your text link analysis (TLA) rules in the Rule lab before applying them to your complete dataset. The Rule lab is an interactive testing environment within the Text Analytics Workbench where you can enter sample text and see how your existing TLA rules match patterns in the sample. When you find patterns that work, you can automatically generate new TLA rules based on the simulation results. With this iterative approach, you can perfect your rules on small samples to save time and improve accuracy before processing large datasets.

Version 13.0.0 of the SPSS Modeler service includes various fixes.

For details, see What's new and changed in SPSS Modeler.

Select a specific model to validate unstructured synthetic data

You can now select a specific LLM to use for the LLM-as-Judge validator model. The validator model validates the unstructured synthetic data that the generation LLM produces through the knowledge data builder. Previously, the model that you picked for generation was automatically used for validation as well, and you could not pick a different model for validation.

Version 13.0.0 of the Synthetic Data Generator service includes various fixes.

For details, see What's new and changed in Synthetic Data Generator.

Process unstructured documents in multiple languages

You can now ingest and curate unstructured data documents in the following languages:

• French
• German
• Italian
• Japanese
• Korean
• Polish
• Spanish

Use semantic chunking in Unstructured Data Integration

You can now select semantic chunking in the Chunking operator. This option produces chunks that follow natural topic and meaning boundaries rather than arbitrary size limits, resulting in more coherent context units, higher‑quality embeddings, more accurate retrieval, and reduced noise during downstream question‑answering.

Summarize chunks with AI in Unstructured Data Integration

Generate AI-powered summaries for each document chunk to improve context understanding and retrieval accuracy.

Ingest and store unstructured data by using more supported connectors

You can now ingest data from the following sources:

• Confluence
• Google Drive

You can also use the following target databases for vector store:

• OpenSearch
• DataStax Astra DB

You can use the following databases for storing document sets and for entity store:

• Microsoft Azure Databricks
• PostgreSQL
• Db2
• Oracle

Unstructured data curation supports a subset of these connectors.

Work with more file types in Unstructured Data Integration

You can now process the following file types:

• HTML
• XLSX
• BMP
• GIF
• JFIF
• JPG
• JPEG
• PNG
• TIFF
• TIF

Unstructured data curation supports a subset of these file types.

Version 5.4.0 of the Unstructured Data Integration service includes various fixes.

For details, see What's new and changed in Unstructured Data Integration.

Version 1.14.0 of the Voice Gateway service includes various fixes.

For details, see What's new and changed in Voice Gateway.

Version 5.4.0 of the Watson Discovery service includes various fixes.

For details, see What's new and changed in Watson Discovery.

New job operator role for deployment spaces

You can now assign the job operator role to collaborators in deployment spaces. This new role is designed for production environments where you need to delegate job execution and scheduling responsibilities without granting full editing privileges.

The new job operator role is available when adding collaborators to deployment spaces. Job operators can manage job runs while maintaining the integrity of job definitions and assets. When a job operator runs a job, the job executes with the privileges of an editor or admin who has job permissions delegated to them in the deployment space.

By using the job operator role, you can achieve the following benefits:

• Enhanced security: Separate job execution responsibilities from asset modification privileges

• Production readiness: Enable operations teams to manage job runs without risking changes to production assets

• Flexible delegation: Allow editors and admins to delegate routine job management tasks while maintaining control over definitions

• Audit compliance: Maintain clear separation of duties for regulated environments

Scale your AI workloads with asynchronous execution

You can now run AI functions and AI services asynchronously to improve performance and scalability for I/O-bound tasks.

When you define async generator functions, the runtime automatically detects and executes them concurrently, enabling external API calls, file I/O, and parallel async tasks.

For AI services, all three methods—generate(), generate_stream(), and generate_batch()—support asynchronous execution for fully async request handling, streaming responses, and batch jobs. The runtime automatically detects whether your function or service is synchronous or asynchronous, so you can adopt async patterns without changing your deployment process.

Deploy ML models with the new onnxruntime_opset_21 software specification

You can now deploy machine learning models with the new onnxruntime_opset_21 software specification. The specification provides enhanced performance and compatibility with the latest ONNX model formats.

Configure runtime-specific ephemeral storage to ensure predictable scheduling and safer capacity planning

You can now configure ephemeral storage requests and limits for each runtime type in AutoAI. By setting these values in the WmlBase custom resource, you prevent runtime pods from being evicted when disk space runs low and ensure that workloads have the temporary storage they need for model downloads, scratch data, and logs.

For more information, see Configuring ephemeral storage for runtime definitions.

Version 5.4.0 of the Watson Machine Learning service includes various fixes.

For details, see What's new and changed in Watson Machine Learning.

Version 5.4.0 of the Watson OpenScale service includes various fixes.

For details, see What's new and changed in Watson OpenScale.

New Dutch and Italian large speech models

You can now use the following large speech models with Watson Speech services:

• Dutch (Netherlands) (nl-NL)
• Italian (Italy) (it-IT)

For details, see Large speech languages and models.

Version 5.4.0 of the Watson Speech to Text service includes various fixes.

For details, see What's new and changed in Watson Speech to Text.

Version 13.0.0 of the Watson Studio service includes various fixes.

For details, see What's new and changed in Watson Studio.

Version 13.0.0 of the Watson Studio Runtimes service includes various fixes.

For details, see What's new and changed in Watson Studio Runtimes.

New foundation models in watsonx.ai

You can now use the following foundation models for inferencing from the Prompt Lab and the API:

• ministral-3-14b-instruct-2512-bf16
• ministral-3b-instruct-2512
• voxtral-mini-2507
• nvidia-nemotron-nano-12b-v2-vl-fp8
• nvidia-nemotron-3-nano-30b-a3b-fp8
• granite-4-1b-speech

For details, see Foundation models.

Generative AI inferencing and tools available on IBM Z and IBM LinuxONE with IBM Spyre

You can now deploy foundation models on IBM Z and LinuxONE systems running the s390x architecture with the IBM Spyre hardware accelerator. Hardware accelerators improve inference performance for AI workloads on mainframe infrastructure.

The following features are available in this release:

• Guardrails
• Prompt Lab
• Watson Document Understanding
  • Semantic Text Extraction
  • Processing API
  • KVP extraction APIs implemented for LinuxONE and Linux on IBM Z

Use semantic schema APIs to extract schema-based key-value pairs

You can now generate custom key‑value pair schemas automatically from sample documents by using Semantic Schema APIs. These APIs reduce manual effort, improve scalability across document types, and enable schema‑based extraction when no suitable predefined schema is available.

Optimize key-value pair extraction by using parameters

You can now improve performance when you extract key-value pairs by using schema-based extraction with specific request parameters. If you know the document type in advance, you can specify the schema by using force_schema_name to skip automatic classification and reduce processing time.

Use model gateway models in Prompt Lab

You can now use models that are configured in the model gateway in Prompt Lab. In addition to IBM foundation models, you can choose models from configured gateway providers, including deployed and custom models. Models must be configured in the model gateway before they are available in Prompt Lab.

Deploy AI services by using the GPU-based cuda runtime

You can now deploy AI services by using the new runtime-25.1-py3.12-cuda software specification. Use the new software specification for resource-intensive computations, for example, to facilitate running embedding models.

Configure and manage the model gateway directly from the watsonx.ai UI

You can now use the watsonx.ai UI to configure and manage model gateway. Edit existing connections and models, enable load balancing to distribute traffic efficiently across multiple models, create access policies to define which groups can access specific resources, and set rate limits to control and prevent request overload.

Access a broader range of foundation models from the model gateway interface

You can now configure additional model providers in the model gateway, expanding your options for accessing foundation models. Use the model gateway to access Cohere, Groq, Mistral, Ollama, and xAI models.

Specify model modality in the UI when you import custom foundation models

You can now specify the task (modality) of a custom foundation model when you import it into your deployment space by using the UI. Example task modalities include audio chat, image chat, and text chat. The available modalities depend on the configuration settings that were defined by administrators when the custom foundation model was added in IBM Software Hub.

Reference subfolders in AutoAI RAG experiments for more flexibility

You can now reference files in subdirectories when you create AutoAI RAG experiments. All file IDs must use absolute paths from the bucket root (for example, data/subfolder/document.txt).

When you reference files in subdirectories, you have more flexibility in organizing your document collections for RAG experiments. If you don't use subfolders, and all files are in the same path, you can refer to them by just their filenames.

You do not have to specify paths for project data assets.

Configure runtime-specific ephemeral storage to ensure predictable scheduling and safer capacity planning

You can now configure ephemeral storage requests and limits for each runtime type in AutoAI. By setting these values in the WmlBase custom resource, you prevent runtime pods from being evicted when disk space runs low and ensure that workloads have the temporary storage they need for model downloads, scratch data, and logs.

For more information, see Configuring ephemeral storage for runtime definitions.

Version 13.0.0 of the watsonx.ai service includes various fixes.

For details, see What's new and changed in watsonx.ai.

Version 5.4.0 of the watsonx Assistant service includes various security fixes.

For details, see What's new and changed in watsonx Assistant.

AWS Databricks connector

You can now connect to data in AWS Databricks to create metrics and ask questions in watsonx BI.

For more information, see AWS Databricks connection.

Microsoft Azure Databricks connnector

You can now connect to data in Microsoft Azure Databricks to create metrics and ask questions in watsonx BI.

For more information, see Microsoft Azure Databricks connection.

Ask questions that use multiple assets

You can now ask natural‑language questions that use multiple data assets in a single query. In earlier versions, watsonx BI selected one best‑fit asset to answer a question. With multi‑asset queries, watsonx BI can retrieve data from one asset, use that result to build a new query, and return a combined answer. With this approach, you can ask multi‑step questions such as questions about comparisons and rank‑based filters.

For example, you can ask:

• For the product with the most revenue, how many returns did it have?
• What is the return count for the top 10 SKUs by revenue in 2025?

Watsonx BI automatically determines which assets are required, issues independent queries, and generates a combined answer. The multi-asset query capability works when chain-of-thought reasoning is enabled in watsonx BI and all referenced assets are stored in the same container, such as a project or catalog.

For more information, see Asking questions across multiple data assets.

Chain-of-thought reasoning

Chain-of-thought reasoning is now available with OpenAI gpt-oss-120b. When you enable chain-of-thought reasoning, you can see step-by-step reasoning for complex or multi-part questions. which provides structured answers and visibility into how queries are built. Reasoning might include chosen data sources, intermediate calculations, and generated SQL.

For more information, see Chain of thought.

Version 3.5.0 of the watsonx BI service includes various security fixes.

For details, see What's new and changed in watsonx BI.

Version 5.4.0 of the watsonx Code Assistant service includes various security fixes.

For details, see What's new and changed in watsonx Code Assistant.

Version 5.4.0 of the watsonx Code Assistant for Red Hat Ansible Lightspeed service includes various fixes.

For details, see What's new and changed in watsonx Code Assistant for Red Hat Ansible Lightspeed.

Track the usage metrics using UMS (Usage Metering Service)

You can now track the usage metrics for IBM watsonx Code Assistant for Z using UMS (Usage Metering Service). It helps you to monitor product usage, ensure license compliance, and gain deeper insights into how the tool is being adopted. UMS is enabled by default.

Gain real-time visibility into applications using IBM Instana

You can now gain real-time visibility into your applications, infrastructure, and dependencies using IBM Instana. It helps you to understand what’s happening inside your systems by examining the telemetry data, such as logs, metrics, traces or API flows.

By default, IBM Instana is disabled. To enable Instana metric collection, run the following patch command:

oc patch wcazagents wcazagents-cr -n <NAMESPACE> --type='json' -p='[{"op": "replace", "path": "/spec/enableInstanaMetricCollection", "value": true}]'

Replace <NAMESPACE> with the namespace where your agents are deployed.

Version 2.8.0 of the watsonx Code Assistant for Z Agentic service includes various fixes.

For details, see What's new and changed in watsonx Code Assistant for Z Agentic.

Deploy Z Understand

You can now easily deploy and configure full IBM Z Understand and Refactor solution on IBM Software Hub. For more information see, Deploying Z Understand on IBM Software Hub.

Shutdown, backup and restore to a different namespace

You can now shutdown, backup, and restore to a different namespace.

Business rules analysis view

You can now view the business rule documents using the Business Rules Analysis view.

Synchronize mainframe members

You can now define how Z Understand containers synchronize members from specific mainframe libraries, including adding or removing members in project source files. For more information, see Synchronize Mainframe Members.

Automated text replacement in source files

By applying regular expression–based search-and-replace rules during preprocessing, you can now standardize code patterns, ignore unwanted constructs, or temporarily modify code structure before analysis. For more information, see Building Decisions.

Conditional compilation in COBOL

COBOL compilation directives are special instructions evaluated by the compiler during compilation (not at runtime). They control how the source code is handled and determine which sections of code are included or excluded in the final program. For more information, see Conditional compilation directives in COBOL.

API Extensibility

You can now access an analysis reflecting usage of in-house or third-party APIs using a configuration file. JSON configuration files describe how each API/Macro call is interpreted, eliminating the need to wait for development support. For more information, see API Call/Macro Extensibility.

Container sync, TLS, and Codepage configuration for PDS libraries

Administrators can now manage mainframe router credentials for z/OS® systems. The Build Client Install Configuration provides a centralized user interface to create, view, edit, update, and delete z/OS connection configurations. For more information, see Build Client Install Configuration.

Comprehensive DDL scripts

You can now use comprehensive DDL scripts for audit and cross application databases.

Version 2.8.30 of the watsonx Code Assistant for Z Understand service includes various fixes.

For details, see What's new and changed in watsonx Code Assistant for Z Understand.

Version 2.4.0 of the watsonx.data service includes various fixes.

For details, see What's new and changed in watsonx.data.

Version 2.4.0 of the watsonx.data Premium service includes various fixes.

For details, see What's new and changed in watsonx.data Premium.

Connect to AlloyDB for PostgreSQL databases

You can now use the AlloyDB for PostgreSQL connector in your DataStage flows to read and write data from AlloyDB for PostgreSQL databases.

Access data in AWS Databricks

You can now use the AWS Databricks connector in your DataStage flows to access and process data in Databricks workspaces.

Access files in Microsoft SharePoint

You can now use the Microsoft SharePoint Files on Canvas connector in your DataStage flows to read and write files stored in SharePoint document libraries.

Access data in Microsoft Dynamics 365

You can now use the Microsoft Dynamics 365 connector in your DataStage flows to read and write business data from Dynamics 365 applications.

Export and import compiled pipeline binaries

You can now export and import compiled Python binaries with optimized runner pipelines, which means that you can move pipelines together with their compiled assets. You control this behavior by using the include-python-binaries and include-common-binaries options in cpdctl.

Data encryption for Teradata connections

You can now enable full session data encryption for Teradata optimized flows by using the new Data Encryption option. This option uses either TDGSS or TLS/SSL to encrypt network traffic, SQL statements, data requests, and responses for the entire session.

Create parameter sets from connection properties

You can now create parameter sets directly from connection properties for supported connectors. Select one or more connection types and add their properties as parameters so that you can easily reuse and manage configuration values across pipelines.

Run remote engines on s390x remote engines

You can now run remote engines on s390x (IBM Z and LinuxONE) systems, deployed as Docker containers or in Kubernetes clusters. This allows you to submit jobs from x86_64 environments and execute them on s390x hardware. This capability enables workload distribution across heterogeneous architectures.

Receive alerts in Microsoft Teams or PagerDuty

You can now create alert receivers to connect Data Observability to your Microsoft Teams channels or PagerDuty services. When you create a PagerDuty alert receiver, you can track triggered alerts and manage events with your existing PagerDuty services. When you create a Microsoft Teams alert receiver, you can receive detailed notifications about triggered alerts in your Microsoft Teams channels.

Identify trends in your data by using metric charts

You can now add metric charts to your Data Observability dashboard. By adding metric charts, you can easily see how a metric has changed across jobs runs, which can help you identify trends in your data.

Reuse connection details in StreamSets flow

When you deploy a Data Collector engine version 7.4.0, you can include connections in StreamSets flows.

Easily manage and reuse StreamSets flows by using parameters

You can now use parameters in StreamSets flows to set values for stage properties at run time. You can change parameter values for each job run without editing the flow, making your flows easier to manage and reuse.

Choose how your browser connects to StreamSets engines

StreamSets engines can now use the tunneling communication method, giving you more flexibility in how your browser connects to the engine. With tunneling, the browser communicates with watsonx.data integration, which securely relays data to the engine through an encrypted tunnel. This method requires no additional setup and is enabled by default.

Run multiple engines for a StreamSets environment to support job failover

When you run multiple engines for a StreamSets environment, jobs can now fail over to another engine if the current engine becomes unavailable. The job restarts on an available engine and continues processing from where it stopped.

Track StreamSets job run history

You can now view a detailed history of a StreamSets job run to diagnose issues and understand the run state, including cases where a run remains in the Queued or Canceling status. The run history lists timestamped events that show status changes, retries, failovers, and other run activities.

Capture a snapshot of data as it moves through a StreamSets job run

You can now capture and view a snapshot to verify how a StreamSets job processes data. A snapshot is a set of data that is captured as it moves through a running job.

Similar to previewing a flow, you can view how snapshot data moves through a job stage by stage. You can drill down to review the values of each record to determine whether the stage transforms data as expected.

Process unstructured documents in multiple languages

You can now ingest and curate unstructured data documents in the following languages:

• French
• German
• Italian
• Japanese
• Korean
• Polish
• Spanish

Use semantic chunking in Unstructured Data Integration

You can now select semantic chunking in the Chunking operator. This option produces chunks that follow natural topic and meaning boundaries rather than arbitrary size limits, resulting in more coherent context units, higher‑quality embeddings, more accurate retrieval, and reduced noise during downstream question‑answering.

Summarize chunks with AI in Unstructured Data Integration

Generate AI-powered summaries for each document chunk to improve context understanding and retrieval accuracy.

Ingest and store unstructured data by using more supported connectors

You can now ingest data from the following sources:

• Confluence
• Google Drive

You can also use the following target databases for vector store:

• OpenSearch
• DataStax Astra DB

You can use the following databases for storing document sets and for entity store:

• Microsoft Azure Databricks
• PostgreSQL
• Db2
• Oracle

Unstructured data curation supports a subset of these connectors.

Work with more file types in Unstructured Data Integration

You can now process the following file types:

• HTML
• XLSX
• BMP
• GIF
• JFIF
• JPG
• JPEG
• PNG
• TIFF
• TIF

Unstructured data curation supports a subset of these file types.

Version 2.4.0 of the watsonx.data integration service includes various fixes.

For details, see What's new and changed in watsonx.data integration.

Run multiple data quality rules with a single click

To run multiple data quality rules at once, you can now organize them in data quality rule groups.

Import, enrich, and assess the quality of data from additional data sources

You can now import metadata from Microsoft Azure Fabric Warehouse data sources, enrich that data, and assess its quality.

Query data assets and document libraries in natural languages other than English

The Text-to-SQL service now uses a multilingual embedding model so that you can create natural language queries in languages other than English. In upgraded deployments, you must reprocess existing metadata with the new model.

Sync reference data sets to external databases

Reference data sets can now be synchronized to external databases and then consumed as regular governed data assets. With this enhancement, a reference data set can be profiled, queried via SQL, joined with other data sets, and reused consistently across the platform. This enables broader consumption of reference data for validation, standardization, analytics, and AI use cases.

Import and export reporting queries

You can now export your reporting queries in batch into a JSON file, and import JSON files with queries. Bulk import and export lets teams easily migrate, back up, and manage large collections of reporting queries across environments.

Main menu improvements

You can now view and manage all artifacts grouped by type, and all categories by opening Governance > All artifacts from the main menu.

Control data and metadata access with the Access data preview permission

You can now assign the Access data preview permission to users and user groups across all projects, catalogs, and deployment spaces to have more control over who can view the actual data and address security requirements. Users without this permission can view asset metadata, but they are blocked from previewing actual data.

By default, the new permission is included in the following predefined roles:

• Administrator
• Business Analyst
• Data Engineer
• Data Scientist
• Data Steward
• Data Quality Analyst
• Developer
• User

If watsonx.ai and DataStage are deployed, the following tools and views are disabled for users and user groups without the permission:

Watsonx.ai

• AutoAI
• SPSS Modeler
• Jupyter
• Pipeline
• RStudio

DataStage

• Data previews (canvas)
• Data previews for data sets and file sets

Update role assignments and any custom roles that you have for users who need to preview asset data.

Browse asset hierarchies

You can now explore and manage connected, custom, and unstructured assets from the Hierarchies pane on the catalog Assets page. Information about relationships between available data sources, databases, schemas, tables, and columns in catalogs is available in one structured view. With these details, you can, for example, navigate relationships across multiple asset levels, view detailed information for assets at any level, or bulk select assets across levels to manage metadata more efficiently.

Access data assets with AWS Databricks delivery method

You can now use the AWS Databricks delivery method to access a data product with AWS Databricks assets. After you subscribe to a data product and it is delivered to you, you can directly access the AWS Databricks resources and view details about how to use them.

Access data assets with Snowflake delivery method

You can now use the new Snowflake delivery method to access a data product with Snowflake assets. After you subscribe to a data product and it is delivered to you, you can directly access the Snowflake resources and view details about how to use them.

Integrate notebooks across watsonx.data intelligence and projects

When you install Watson Studio or watsonx.ai, you can use Data Product Hub to share or reuse your watsonx.data intelligence notebooks. As a data producer, you can add notebooks and data assets from your project directly to a data product. When you deliver a data product, recipients can download the included notebooks for their own use. You can also add notebooks to projects from data products.

Customize data product request forms to match your organization's needs

Customize your data product request forms to collect the exact information that your organization requires. Choose and edit multiple field types to create a form that fits your workflow.

Create flexible data product delivery options with custom delivery methods

Create a custom delivery method based on your use case. You can configure your delivery method to be used in restricted or public data products and to support multiple data source types and asset types.

Set up watsonx.data intelligence MCP server and use natural language to interact with data

Now you can do key tasks by using your AI agent and watsonx.data intelligence MCP server. Use natural language prompts to securely access and explore your data and to complete tasks for data governance and catalogs, data quality, data lineage, and Data Product Hub.

Work with Unstructured Data Integration flows in watsonx.data intelligence

You can now install Unstructured Data Integration as an optional component in watsonx.data intelligence to ingest, transform, enrich and curate unstructured data from diverse sources.

Create and manage document classes for unstructured data processing from the UI

You can now create and update document classes that are used in unstructured data curation and integration by using the new document class editor.

Process unstructured documents in multiple languages

You can now ingest and curate unstructured data documents in the following languages:

• French
• German
• Italian
• Japanese
• Korean
• Polish
• Spanish

Use semantic chunking in Unstructured Data Integration

You can now select semantic chunking in the Chunking operator. This option produces chunks that follow natural topic and meaning boundaries rather than arbitrary size limits, resulting in more coherent context units, higher‑quality embeddings, more accurate retrieval, and reduced noise during downstream question‑answering.

Summarize chunks with AI in Unstructured Data Integration

Generate AI-powered summaries for each document chunk to improve context understanding and retrieval accuracy.

Ingest and store unstructured data by using more supported connectors

You can now ingest data from the following sources:

• Confluence
• Google Drive

You can also use the following target databases for vector store:

• OpenSearch
• DataStax Astra DB

You can use the following databases for storing document sets and for entity store:

• Microsoft Azure Databricks
• PostgreSQL
• Db2
• Oracle

Unstructured data curation supports a subset of these connectors.

Work with more file types in Unstructured Data Integration

You can now process the following file types:

• HTML
• XLSX
• BMP
• GIF
• JFIF
• JPG
• JPEG
• PNG
• TIFF
• TIF

Unstructured data curation supports a subset of these file types.

Track data flow changes by comparing lineage versions

You can now compare two versions of a lineage graph to identify which assets were added, removed, or modified between versions. This comparison helps you to understand how your data flows evolved and assess the downstream impact of those changes.

Monitor processed OpenLineage events

You can now monitor OpenLineage events in a centralized dashboard to verify event ingestion, identify failed or pending events, and troubleshoot processing issues. The dashboard also helps you understand the overall health of the OpenLineage processing pipeline by showing event volume and trends over time.

Export data lineage to OpenLineage payloads

You can now export data lineage as an OpenLineage .json payload, making it compatible with any application that supports the OpenLineage standard format.

Import lineage from new data sources

You can now import lineage metadata from the following additional data sources:

• IBM Netezza Performance Server
• Tech preview SAP BusinessObjects
• SAP HANA

Connect to new data sources by using version 1.5.0 of the Manta agent

You can now import lineage metadata from the following data sources by using the updated Manta agent:

• Db2
• Db2 for z/OS
• Db2 on Cloud
• SAP BusinessObjects
• Tableau

Manta agent version 1.5.0 is now available. Version 1.4.0 is deprecated, and versions 1.0.0, 1.1.0, and 1.2.0 are no longer supported. Update unsupported agent versions and consider upgrading version 1.4.0 instances to the latest version 1.5.0.

Pause and resume the import of lineage metadata from Qlik Sense and SAP HANA

You can now pause lineage metadata imports from Qlik Sense and SAP HANA and resume them at your convenience. This control is especially useful when importing large data volumes that require extended processing time.

Delete lineage from a specific point in time

You can now delete lineage data from a specific time range to remove data that you no longer need or that was imported by mistake. Only data imported during your specified time range is deleted, helping you focus on relevant lineage information.

View asset ownership directly on the lineage graph

You can now see the assigned owner, whether a user or a group, displayed directly on assets in the lineage graph. This immediate visibility saves you time by eliminating the need to open the metadata details panel to find ownership information as you explore data lineage relationships.

Version 2.4.0 of the watsonx.data intelligence service includes various fixes.

For details, see What's new and changed in watsonx.data intelligence.

Evaluate detached prompt templates without creating deployments

Detached prompt templates are now evaluated without any dependency on watsonx.ai. As a result, you now evaluate detached prompt templates directly in the deployment space, without creating deployments.

In a deployment space, you now start an evaluation from the Assets tab. Click the menu for a detached prompt template, and then click Evaluate.

You can no longer create deployments for detached prompt templates.

Your existing deployments are not impacted by these changes.

Watsonx.governance is now available on IBM Z and LinuxONE hardware

You can now install the watsonx.governance service on Red Hat OpenShift Container Platform on IBM Z and LinuxONE (s390x) hardware.

Create custom detectors in Guardrail Manager

You can now create custom guardrail detectors in Guardrail Manager. You create a custom guardrail detector by configuring the detector properties, input parameters, and the structure of the requests and responses for the detector.

Previously, you could create custom detectors only through the API.

For more information about new features and other updates in Governance console, see New features in version 9.2.0 in the OpenPages documentation.

Version 2.4.0 of the watsonx.governance service includes various fixes.

For details, see What's new and changed in watsonx.governance.

Handle errors in agentic workflows

You can add error handling to manage tool failures within your agentic workflows. When a tool fails, you can configure the workflow to automatically retry the operation. If the retry attempts fail, the workflow can display a custom error message to the user or continue along an alternative failure path based on your configuration. Now you can:

• Add error branches to any node in your workflow.
• Configure automatic retry logic with customizable retry attempts.
• Define alternative processing paths that execute when errors occur.
• Display custom error messages to users when retry attempts fail.
• Loop back to the original node after fixing issues.
• View error handling paths in flow inspector and observability traces.
• Maintain partial results when errors occur in multi-step processes.

Use this feature to build resilient workflows that can handle API failures, timeout scenarios, and data validation errors without losing progress or requiring manual intervention. For details, see Adding tools.

View flow run data on the canvas

You can open the flow inspector directly from the flow builder by clicking the Open flow inspector icon. Access workflow execution history and inspect how a workflow ran from start to finish without leaving the flow builder interface. Now you can:

• Click the Open flow inspector icon in the flow builder to access execution history.
• View the list of workflow runs with status, initiator, duration, and environment details.
• Filter workflow runs by time range, status, environment, agent, or initiator.
• Select a specific flow run to view detailed execution information.
• Review flow events showing the sequence of actions taken during the run.
• Inspect flow parameters including inputs, outputs, and flow variables.
• Open nodes directly in the flow builder from the flow run details.

Use this feature to quickly diagnose flow issues, review execution patterns, and understand workflow behavior directly from the flow builder interface. For details, see Inspecting agentic workflows.

Control builder access to embedded security settings

As an administrator you can now control whether builders can modify embedded security settings, ensuring only authorized personnel manage security configurations. By default, builders have access to configure security settings. Now you can:

• Toggle the Allow Builders to manage security settings option in the Embedded Security page.
• Restrict builders to read-only access of the Embedded Security page.
• Prevent builders from toggling the Security switch or modifying the public key field.
• Block API requests from builders to embedded security endpoints (returns 403 Forbidden errors).
• Maintain full administrator access to all security settings regardless of builder access setting.
• Apply changes immediately to all builders in your tenant.

Use this feature to enforce security governance policies and ensure only authorized personnel can modify embedded chat security settings. Changes take effect immediately and can be reversed at any time. For details, see Restricting builder access to security settings.

Work with time-based data in agentic workflows

You can handle time-based information in your agentic workflows with new Date/Time and Date/Time range fields, making it easier to schedule appointments, track work hours, and manage time-sensitive processes. Capture and process time data with time zone support. Now you can:

• Define time variables that store time values (hours and minutes with optional timezone).
• Use the date and time data type to store both date and time information in a single variable.
• Select from a comprehensive list of time zones, which is displayed with UTC offset and city name.
• Let users select time values by using intuitive time picker widgets in chat.
• Handle time ranges by defining start and end times for scheduling scenarios.

For details, see Date/Time, Date/Time range and Input, output, and variable types.

Control model selection visibility

You can control whether builders see model selection options in the UI. By default, model selection is hidden for new tenants, simplifying the builder experience for organizations that prefer to use default model configurations. Now you can:

• Turn model selection on or off from the Settings page.
• Hide model selection for agents while keeping it available for gen AI nodes in agentic workflows.
• Maintain existing model selection settings for current tenants.
• Enable model selection when builders need to choose specific models.

Use this feature to standardize on specific models while reducing complexity for builders who don't need to make model choices. For details, see Managing model selection settings.

Upload multiple files to tools

You can upload multiple files simultaneously to tools with configurable file type restrictions, size limits, and validation rules up to 30 MB per file. Now you can:

• Upload up to 100 files in a single operation.
• Configure maximum file size per file (up to 30 MB) and total upload size.
• Define allowed file types such as documents, spreadsheets, images, audio, and code files.
• Use the new WXOFile class for efficient file handling with lazy loading.
• Customize the file upload prompt message.

Customize tool fields for domain agents

You can customize tool fields for domain agents directly in the UI to accommodate client-specific field names and custom implementations without requiring code changes. Now you can:

• Add, remove, or modify tool field names during agent configuration.
• Adapt tools to match client-specific field naming conventions, for example, "ID" versus "IBMid".
• Configure tools for custom Salesforce, ServiceNow, and other integrations.
• Reduce tool errors caused by field name mismatches.

For details, see Customizing the schema for tools.

Use a schema from a list of predefined schemas

You can select from a list of predefined schemas when configuring structured document extractors in your agentic workflows. Simplify configuration by using ready-to-use schemas for common document types such as bank statements, invoices, and insurance claims. Now you can:

• Choose from a curated list of predefined schemas for common document types.
• Automatically add multiple fields associated with the selected document type.
• Reduce manual effort by avoiding the need to add fields one at a time.
• Quickly configure document extractors without manually defining field structures.

For details, see Use a schema from a list of predefined schemas.

Voice security for embedded chat

Voice capabilities in embedded chat now support both authenticated and anonymous modes. When security is enabled, voice audio streams use JWT-based authentication, binding each session to the user’s identity. For details, see Enabling voice capabilities in the embedded agent.

Version 8.0.0 of the watsonx Orchestrate service includes various fixes.

For details, see What's new and changed in watsonx Orchestrate.

• Version 4.16.4 or later fixes
• Version 4.18.6 or later fixes
• Version 4.19.0 or later fixes
• Version 4.20.0 or later fixes
• Version 4.21.0 or later fixes

Many services in IBM Software Hub use Helm for installation and upgrade. For services that support Helm, a cluster administrator must create cluster-scoped resources, such as custom resource definitions, cluster roles, cluster role binding, and webhooks. This change gives cluster administrators more insight into the cluster-scoped resources that are required for each service.

The install-components command replaces the apply-olm and apply-cr commands. The command ensures that the required operators are created before the custom resources are created.

IBM Software Hub Version 5.4 uses the Red Hat OpenShift Container Platform cert-manager Operator.

If you are upgrading to IBM Software Hub Version 5.4, ensure that you migrate to the Red Hat OpenShift Container Platform cert-manager Operator before you upgrade IBM Software Hub.

For details, see Migrating from the IBM Certificate manager to the Red Hat OpenShift Certificate manager

The cpd-cli manage setup-instance command is deprecated. Use the new cpd-cli manage install-components command to install and upgrade the IBM Software Hub platform.

For details, see:

Installation

Installing the required components for an instance of IBM Software Hub

Upgrade from 5.2

Upgrading IBM Software Hub

Upgrade from 5.3

Upgrading IBM Software Hub

The apply-olm command is replaced by the install-components command.

The apply-cr command is replaced by the install-components command.