Explore the Agent's Data Using the Ask About Panel

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.

In a agent, you interact with the data that the agent designerincluded. However, if you are not familiar with that data, you may find it challenging to know what to ask about, beyond the suggestions and prompts designed by the agent designer. Beginning in Strategy One (July 2025), open the Ask About panel to display the agent's content. The Ask About panel lists the available columns and their descriptions (if available). You can start a conversation about any of the columns. The agent response describes the data and, for an attribute, lists the available elements (for example, the actual names of customers). Additional details and insights are provided to help you understand the data so that you can create thoughtful analytic questions. In the example below, the user asked about the Cost metric. More examples are provided in the steps below.

Object descriptions display if that information is available in the dataset, for example, if descriptions were added when data was imported. In the following example, the Cost metric contains a description: The cost spent on each category over the time, across 50 different regions.

This description is separate from the automatically-generated descriptions created when a dataset is enabled for AI. Those descriptions provide the context needed for the agent to make decisions, rather than the context needed to understand the data and ask questions about it.

Use the Ask About panel

The agent's designer must have enabled the Ask About panel for it to be available.

In Library, click a agent to open it. You can use the Filters panel to search for a specific agent name or filter to display only agents, as shown below:
Click Ask About to display the Ask About panel on the left. If the Chats panel is displayed, the Ask About panel replaces it.
A list of the agent's data objects displays so you know what data is available for you to ask about.
If an object contains a description or column alias, display it by expanding the object. These explanations can help you decide what to ask about. Starting in Strategy One (August 2025), any column aliases created for the object are included.
Hover your cursor over the object and click Start Conversation .
The response to starting a conversation about the City attribute is shown below. The question is interpreted as: Show a summary of the column: City. The response describes the City column and provides some examples of the cities in the data. A grid of all the available cities displays. Finally, insights are provided about the object's diverse coverage, strategic locations, and urban vs. rural dynamics.
The response to starting a conversation about the Cost metric is shown below. The question is interpreted as: Show a summary of the column: Cost. The response provides the total, average, minimum, and maximum costs. Insights are provided about the subcategory with the highest cost and the wide range between the minimum and maximum costs. Suggestions to understand the average cost and analyze cost distribution complete the insights.
The response to starting a conversation about unstructured data, in this case a PDF, is shown below. The response describes the report and provides insights about customer data platforms driving loyalty, omnichannel boosting spending, inventory accuracy improving with real-time data, and AI recommendations increasing conversions.
You can start a conversation about any of the source agenta in a universal agent. The response to starting a conversation about the Airlines Bot is shown below. The agent is described, including the kinds of queries it supports. Insights are provided about on-time rates, optimal departure windows, short-haul inefficiency, and global reach.
You can work with the Ask About response just as with any other agent response. For example, you can:
- Choose one of the related suggestions about the object or, based on the information and insights provided, ask a question of your own.
- View AI Interpretation for insights on how the agent interpreted and executed your question.
- Take a snapshot.
- Export the response.
For steps, see Ask Questions in an Agent.