Use Slate AI as a drafting assistant when you need to create or refine the HTML for a custom dashboard. You can start with a plain-language prompt, preview the generated layout, and then continue prompting until the dashboard presents the record data in the right order and format.
This approach is useful when you know what information the dashboard should show but want help creating the initial layout, styling, and structure.
📝 Note
This article describes using Slate AI to help edit a standard custom dashboard. This is different from the separate AI Dashboards feature, which uses configured query bases and prompts to analyze record data.
⭐ Get Inspired
This article was adapted from a post by Technolutions staff in the Slate Community Forums' Get Inspired space. Have a great idea for a Get Inspired post? Let us know!
Before you begin
Before you prompt Slate AI, decide which record values the dashboard should show. A clear list of fields helps Slate AI produce a more useful first draft.
For example, an admissions-focused dashboard might include:
Application major
Student type
Entry term
Application status
Current school and GPA
Sports or interests
Assigned counselor
If the dashboard uses query exports as merge fields, give the exports clear names that are easy to reference in the dashboard editor. Avoid names with unnecessary punctuation or spacing that could make merge fields harder to reuse.
Create a first dashboard draft
Open the dashboard editor for the record dashboard you want to build.
Start with an empty dashboard or a minimal existing layout.
Open Slate AI from the editor.
Enter a prompt that describes the dashboard content and the user goal.
Example prompt:
Please create a custom dashboard that highlights the record's application major, student type, and term, their current school and GPA, and the admission counselor assigned to the record.The first response may be structurally correct but visually plain. Treat the first version as a starting point.
Refine the layout with follow-up prompts
Use follow-up prompts to specify the layout, visual treatment, alignment, and actions that should appear in the dashboard. Be explicit about what should change and what should stay in place.
Example follow-up prompt:
Please update the dashboard to be more visually appealing with a modern and sleek design, and adjust the formatting so that major, student type, and term are on one line, school and GPA are on another, and the assigned counselor is on its own line. Also, please add a button labeled "New Interaction." Please left justify the dashboard rather than centering it.
If the result is too narrow, too sparse, or missing information, continue with another prompt.
Example follow-up prompt:
Please make the dashboard full width and add application status to the line with major, student type, and term. Please also add sports and interests to the line with school and GPA.
Preview the dashboard on a record
After the layout is close to final, preview the dashboard on an actual record. This check confirms that merge fields populate as expected and that the design still works when real values are longer, shorter, or blank.
Review the generated HTML
Slate AI can speed up dashboard drafting, but you should still review the generated HTML before relying on it in production.
Remove unused markup, placeholder text, or sample values.
Confirm that links and buttons point to the correct locations.
Confirm that merge fields match the dashboard query exports.
Preview records with missing values, long values, and different lifecycle statuses.
Keep the dashboard focused on the information that users need during the record review task.