Reuse maintained definitions
The rebuild flow checks existing metrics before creating missing ones. Similar names alone do not prove the definitions match.
Datus + Apache Superset
Apache Superset AI Agent Integration
Turn the SQL and business logic behind your Superset dashboards into reusable metrics and Agent context. Ask follow-up questions, build a Datus dashboard, or create an HTML report from the knowledge your team already has.
Superset dashboardsSelected SQL + chart intent
Reviewed metrics + Reference SQL
Datus agents, reports + dashboards
From dashboard to shared context
Bring the definitions behind the charts with you.
Select the dashboard and charts, confirm the scope, then let Datus build from the exported query evidence. Existing metrics are reused when they match; missing metrics can be authored and validated for the selected scope.
Start from the actual Superset dashboard.
The plugin resolves the dashboard, charts, datasets, and underlying database connection. Your team confirms which queries become Reference SQL and which supply metric evidence.
- Dashboard + chart inventory
- Physical datasource match
- Confirmed query selection
Superset → Datus
Superset dashboard ├─ selected chart queries ├─ dataset / database identity └─ measures, filters, time grain
Explanatory workflow based on the plugin and Agent source. Source filters, permissions, and unsupported chart behavior must be checked; visual similarity alone does not establish equivalent calculations.
Follow the dashboard bootstrap tutorialThree destinations, different jobs
Keep the dashboard. Extend what you can do with it.
Choose where the result belongs. A Superset dashboard, a Datus HTML report, and a Datus HTML dashboard have different interaction and hosting needs.
Lives in Superset
Create and maintain assets in Superset.
Ask Datus Agent to use the Superset authoring plugin to inspect or create datasets, charts, and dashboard layouts. The agent reads assets back and checks representative queries.
Uses your Superset deployment and data connections.
Read the Superset authoring walkthroughApache Superset
By region
Monthly trend
Illustrative layouts; no live dataset or generated customer results are shown.
Why bring BI knowledge into Datus?
Use the work behind one dashboard across more questions.
Carry the source evidence
Keep chart-to-query references and exported SQL with the generated context, so the team can review where a calculation came from.
Choose the right surface
Leave ongoing BI reporting in Superset, use a report for a briefing, or build a Datus dashboard for a new exploration workflow.
Connect your environment
Connect Superset. Confirm the source. Start with one dashboard.
Open the setup and bootstrap guide- Install and enable the Superset plugin with an authorized profile.
- Configure the physical database used by the selected Superset datasets as a Datus datasource.
- Choose the dashboard, confirm the import scope, and review generated context before wider use.
The plugin supports Apache Superset 4.x and later. Authentication, source identity, and per-chart coverage are checked for the selected deployment.
FAQ
Apache Superset integration questions
Can Datus turn Superset dashboard logic into metrics?
Datus can export selected chart queries and use them as evidence for Reference SQL and semantic modeling. The dashboard bootstrap workflow requires a confirmed scope and a matching physical datasource. For a Datus dashboard target, compatible existing metrics are reused and missing selected metrics can be authored. Review the calculation, filters, and time grain rather than assuming the chart title defines the metric.
Can I convert a Superset dashboard into a Datus dashboard?
The current plugin includes guidance for this target, coordinated by the shared dashboard-bootstrap workflow. It resolves the source, exports selected SQL, prepares the necessary metrics, and hands the evidence to gen_visual_dashboard. Coverage is checked per chart. Filters, permissions, unsupported calculations, and requested layout differences need explicit validation before treating the destination as equivalent.
How is a Datus report different from a Datus dashboard?
A Datus HTML report is a narrative snapshot with results captured during generation. A Datus HTML dashboard is an interactive data view whose supported filters send queries to a configured backend. In the CLI workflow that backend is served by datus --web. Both can use metrics and SQL, but they serve different reading and exploration tasks.
Does importing context modify my Superset dashboard?
The Datus-target reconstruction workflow keeps source BI assets read-only apart from exporting selected evidence into local files. Creating or changing a dashboard inside Superset uses the separate Superset authoring workflow, where writes require confirmation. Choose the output destination explicitly so the agent applies the appropriate source and write boundaries.
Where do the data queries run?
Queries use the configured datasource and its access permissions. Superset source discovery resolves the actual database behind a dataset; the Superset server address is not itself the warehouse connection. Datus dashboard filters need the configured query backend. Generated files alone do not replace that backend or prove that the source and destination queries produce equivalent results.
Put your existing BI knowledge to work.
Start with one Superset dashboard and a question your team wants to answer.
Get Started Free