Board 15.1 improves how Developers create, describe, and manage Datasets. The release adds separate Dataset types, type-specific limits, and Data Attribute descriptions that provide clearer business context for users and Board AI Agents.
For an overview of how Datasets provide a semantic data layer for supported Board features, see About Datasets.
Key highlights
Choose between an “AI Dataset” and a “Standard Dataset”.
Create up to 100 Datasets of each type.
Add business-friendly descriptions to Data Attributes.
Reuse suggested descriptions for the same Entity or Cube from other Datasets.
Follow dedicated configuration guidance when preparing Datasets for Board AI Agents.
Choose a Dataset type
When creating a Dataset, Developers can now choose one of the following types:
AI Dataset. Use this Dataset type with Board AI Agents.
Standard Dataset. Use this Dataset type for other supported Board features.
Developers can create up to 100 AI Datasets and 100 standard Datasets in each Data Model. Counters in the Dataset Catalog and the Dataset configuration panel show the number of Datasets in use for each type.
When a Dataset type reaches its limit, Board prevents the creation of more Datasets of that type. Delete or merge a Dataset that is no longer needed before creating another one.
For steps to create an AI Dataset or a Standard Dataset, see Create and configure a Dataset.
Describe Data Attributes
The new “Data Attributes Table” appears after a Developer configures the Dataset Layout in the “Data Configuration” section.
The table lists the elements used in the Layout, including Algorithms, Entities, Cubes, Rules, and Rankings. It also shows where each Data Attribute is used in the Layout.
Developers can add a plain-language description of up to 1,000 characters to each Data Attribute. These descriptions help users understand business-specific or technical terms. They can also help Board AI Agents interpret ambiguous names, abbreviations, and key performance indicators with the correct business context.
A Developer can reuse a suggested description when the same Entity or Cube is already described in another Dataset. This helps keep business definitions consistent across Datasets.
For steps to add, edit, or reuse a Data Attribute description, see Create and configure a Dataset.
Configure AI Datasets
Clear Dataset metadata and a well-structured Quick Layout help Board AI Agents select and interpret data correctly.
Use business-friendly Dataset names and descriptions. Configure clear Block headers, comparable levels of detail, relevant axes, and scope Selections. Include only the Entities and data required for the intended use cases.
For guidance on Dataset metadata, Block configuration, Quick Layout axes, Selections, and governed Entities, see Best practices for AI Datasets.