Demos
Feature walkthroughs by the team building Cube
A chat panel docked on any published or embedded dashboard — ask questions in plain language, answered with live queries against your model.
Cube's MCP server now exposes multiple Cube agents, letting the MCP client pick the right one based on each question's context.
Let the Cube AI agent run real Python for forecasting, statistical tests, and modeling — going beyond what SQL aggregation alone can do.
Set the dashboards, workbooks, and folders embedded users see on first login, so they start with useful content instead of an empty workspace.
Map charts now support regions, plus a round of refinements — visualize geographic data with more control inside workbooks.
Query your semantic model directly in SQL — reference measures and dimensions by name instead of rebuilding the logic in raw SQL.
Audit every access policy in your data model — row-level filters, member restrictions, and masking — broken down by user group.
Share an Analytics Chat thread with teammates as a read-only conversation — per-user, per-group, or org-wide access.
Creator Mode users can now share workbooks with teammates from inside your embedded app — per-user or group access.
A deep dive into Cube's Table chart — column controls, conditional formatting, text and background styling, sorting, and pivots.
Organize semantic views into named collections by domain — helping agents, embedded apps, and people find the right dataset.
