Strategy ONE
Enable a Dataset for Agents
Beginning in Strategy One (September 2025), enhanced bots are available out of the box and have been renamed to agents to reflect the improvements and advancements.
-
For details on the 2.0 enhancements, see Agent Enhancements.
Starting in Strategy One (March 2025), before you can use a dataset to create an Agent, you must enable it for AI. This pre-processing ensures seamless integration and processing of relevant data for AI functionality like natural language querying, predictive analytics, and advanced insights. The cube is enriched with automatically-generated descriptions of the cube itself and each column, providing the context needed for your agents. Enabling a dataset for AI ensures that the data is well-organized, detailed, and ready to answer your agent's questions.
-
Beginning in Strategy One (June 2025), you can add unstructured data files as data sources for agents. Unstructured data includes PDF, Microsoft Word, HTML, markdown, and text files.
-
Beginning in Strategy One (August 2025), you can define custom tags and assign them to unstructured data. Tags act as descriptive labels and improve search results. For steps, see Add Tags to Unstructured Data.
Because you enable only the datasets that are relevant for your agents, you avoid overloading the system by processing all datasets. This also ensures your users buildagents on AI-ready datasets that are relevant, secure, and come from a trusted source.
Prerequisites
-
You must have the Administer Cubes privilege.
-
Cognitive Search must be running, which helps Workstation identify which datasets meet the requirements lists below.
-
The dataset must be an MTDI cube, an In-memory OLAP cube, or a subset report without prompts. The MTDI cube can contain a single table or multiple tables. Beginning in Strategy One (June 2025), the dataset can also be unstructured.
-
A subset report is based on a single cube. That cube must be an MTDI cube or OLAP cube.
-
-
The dataset must be published before you can enable it.
Select one of the following procedures:
Enable a Dataset
- Open the Workstation window.
- Connect to an environment with Administrator privileges.
- In the Navigation pane, click Datasets.
- Select the Datasets tab.
- Right-click the dataset and select Enable for AI.
- If you do not see this option, the Prerequisites have not been met.
- If the dataset is not published, a message displays. Click OK then publish the cube before trying again.
During pre-processing, a loading icon
displays. The system ingests the information and prepares it so it can be used to answer questions. Once the process is complete, the dataset displays the Enabled for AI icon
, as shown below:
-
If the dataset could not be enabled, the dataset displays an error icon
. Hover over the error icon
to display a message about the failure.
-
You can disable the dataset for AI, if it should no longer be used for agents. Right-click the dataset and select Disable for AI.
Add Unstructured Data
- Open the Workstation window.
- Connect to an environment with Administrator privileges.
- In the Navigation pane, click Datasets.
- Select the Unstructured Data tab.
- Click Add Unstructured Data at the top of the pane.
- Select the Project to add the dataset to.
- Drag and drop the file or click Browse Files to locate and select the file.
- Click Add.
During pre-processing, a loading icon
displays. The system ingests the information and prepares it so it can be used to answer questions. Once the process is complete, the file displays a green check
.
Add Tags to Unstructured Data
After you add an unstructured data file to your project, define custom tags and assign them to the file. Tags act as descriptive labels, similar to hashtags on social media, that help to identify, categorize, and group your data based on relevant attributes or themes. This improves search results so that agents can more easily locate related items and perform targeted queries within large collections of unstructured data.
- Open the Workstation window.
- Connect to an environment with Administrator privileges.
- In the Navigation pane, click Datasets.
- Select the Unstructured Data tab.
- Right-click the file and select Properties.
- In the navigation bar on the left, click Tags.
- Click Add Tag.
- Type a Key and then at least one Value for the key. Separate multiple values with commas.
The key is the label that provides context and grouping, such as Country. The values are the specific elements of that key. For a Country key, you can add the following values: Afghanistan, Albania, Algeria, Andorra, Angola, and so on.
- Add more tags as needed.
- To edit a Key or Value, click it and make your changes.
- Click OK to save the tags and return to Workstation.