The BI tool that maintains itself
The fastest data teams don't maintain their BI by hand.
Other tools make maintenance your problem. Lightdash makes it the platform’s.
Proposes changes from real usage; you approve
Manual — your team does it
No governed metrics at all
Autopilot cleans and repairs on a schedule
Manual audits, if they happen
Not handled
Every change reviewed, logged, revertible
Varies by tool
No governance, no audit trail
For any other questions, reach out to our team.
How does the self-improving context layer work?
As your team asks questions, Lightdash spots gaps, outdated definitions, and duplicate metrics in your semantic layer. It proposes each fix as a pull request, validated with a compile. You review the change and approve, reject, or revert it.
What does Autopilot do?
Autopilot is a scheduled agent that keeps your Lightdash project clean. On a cadence you set, it soft-deletes charts with zero views older than 30 days, repairs charts that reference missing fields, flags slow queries, and drafts new charts for questions your team keeps asking. Every action is logged with a reason.
Will letting AI edit my context layer or delete dashboards break things?
No. Nothing changes on its own. Context layer edits arrive as pull requests you review before merging. Autopilot only soft-deletes (everything is restorable), never touches content under 30 days old, and reverses its own mistakes on the next run. You can also run it in read-only mode so it only flags.
How is this different from pointing an LLM at my warehouse?
A raw LLM on your warehouse has no governed metrics, no business context, and no permissions, so it guesses. Lightdash works on your governed semantic layer: every metric is defined in code, every change is reviewed, and every action is logged. You get the automation without losing your source of truth.
Do I need dbt to use self-improving analytics?
dbt is the most common setup. Lightdash builds its semantic layer inside your dbt project, or you can define it in Lightdash YAML against your warehouse without dbt.
I have an infrequently asked question.
Great! We're always around to help. You can chat to us through the live chat on this webpage, or talk to us in our Slack community.
