Ways to build a dashboard

You can build a dashboard from the UI, with REST APIs, or with AI. Use the following table to find the approach that fits how you work, then follow the link to get started. To learn what a dashboard is, refer to Kibana dashboards.

Approach	When to use this	What you get
Create a dashboard from the UI	Exploring data interactively and assembling dashboards panel by panel	Saved dashboard
Create dashboards programmatically	Automating deployments, managing dashboards and visualizations as code, CI/CD pipelines	Saved dashboard or visualization
Create dashboards using AI	Generating dashboards from natural language, or building AI tools that create dashboards	Dashboard through chat that you save when ready, or saved dashboard using the API

Once you have a dashboard, you can enhance it:

Add filter controls to let viewers explore the data interactively
Add drilldowns to open another dashboard, a URL, or Discover from a panel
Arrange and resize panels to optimize its layout
Enhance it with AI to improve its text, layout, and chart styling

You can also maintain your dashboards after you create them: organize your collection with search, tags, and favorites, duplicate a dashboard to reuse it as a starting point, import one from another space, instance, or deployment, or share or export it. For repeatable, auditable changes, manage your dashboards as code to version-control their definitions and deploy them across environments.

Requirements

To create or edit dashboards, you need:

Data indexed into Elasticsearch and a data view. A data view is a subset of your Elasticsearch data, and allows you to load the right data when building a visualization or exploring it.

Tip

If you don't have data at hand and still want to explore dashboards, you can import one of the sample data sets available.
Sufficient privileges for the Dashboard feature. Without them, you might get a read-only indicator. A Kibana administrator can grant you the required privileges.