AI-assisted view configuration helps you create and modify views using natural language, eliminating the need to manually configure every setting.
The first capability available in Beta is AI-assisted Conditional Formatting.
Simply describe what you want to highlight, and the assistant will create the appropriate conditional formatting rules for your current view.
This is the first step toward a broader vision. Over time, the assistant will be able to configure additional view settings, including labels, layouts, simple grouping cases, and more, so you can focus on answering business questions rather than configuring visualizations.
How it works
While you're working in a view, open Omnipresent AI assistant.
The assistant prioritizes the context of your current view. This allows it to understand the data, components, and relationships that are already visible and make changes directly to that visualization.
For example, you can ask:
Highlight applications nearing end of life.
Color business-critical applications in red.
Highlight systems owned by the Finance department.
Highlight applications with no assigned owner.
Dim applications that have an assigned owner.
Emphasize applications with high technical debt.
The assistant interprets your request and generates the corresponding conditional formatting rules automatically.
Tip: To apply conditional formatting, describe how you want the visualization to change using actions such as highlight, color, dim, emphasize, or heat-map. Questions such as "What applications have no assigned owner?" are interpreted as requests for information rather than changes to the current view.
Understanding view context
The assistant behaves differently depending on whether you're asking it to modify the current view or answer a question about your data.
Modify the current view
For example:
Prompt
Highlight applications nearing end of life.
Result
The assistant updates the current visualization by creating conditional formatting rules that highlight those applications.
Ask a question about your organization's data
Prompt
What applications are nearing end of life?
Result
The assistant interprets this as a request for information. Instead of modifying the current view, it searches the available data sources and returns the answer in chat.
In general
Requests that describe how you want the current view to change are applied to the active visualization.
Requests that ask for information are answered in chat using the most appropriate data source.
If your request could reasonably be interpreted in more than one way, the assistant may ask you to clarify your intent before taking action.
How your prompt is interpreted
The examples below illustrate the difference between requesting information and modifying a visualization.
If you ask... | The assistant will... |
Show applications with no assigned owner. | Return a report or answer in chat listing applications with no assigned owner. |
Highlight applications with no assigned owner. | Apply conditional formatting to highlight those applications in the current view. |
Show business-critical applications. | Return information about your business-critical applications. |
Color business-critical applications in red. | Color those applications in the current view. |
What applications are nearing end of life? | Answer the question in chat. |
Highlight applications nearing end of life. | Highlight those applications in the current view |
Example:
What applications are nearing end of life? - will return a table with the list of the "Phasing Out" apps with the Live End Date.
Highlight applications nearing end of life. - will apply conditional formatting in the view.
Beta limitations
During Beta, AI-assisted view configuration supports Conditional Formatting only.
Support for additional view settings is planned for future releases, including:
Labels
Layout configuration
Simple grouping cases
Other visualization settings
We'd love your feedback
This feature is under active development, and your feedback will directly influence what we build next.
If you have suggestions, encounter issues, or have ideas for additional AI-assisted view capabilities, we'd love to hear from you.
Share your feedback on Productboard: LINK.