- Maps your stack. Elementary builds a lineage map of how data moves and the dependencies between assets, with tests, incidents, usage, and ownership attached to it.
- Tests and fixes data in context. Run checks in the pipeline and in the cloud, group related failures into one incident, and trace impact and root cause through lineage.
- Tells agents what to trust. When an agent uses company data, Elementary traces the answer back to the data behind it and flags the issues that affect it, so the result can be trusted, caveated, or withheld.
How it works
You connect your data warehouse to get started. Elementary reads warehouse metadata and runs tests directly against the warehouse. It integrates natively across the modern data stack, from the warehouse to BI, AI tools, and ticketing. From there, Elementary builds a connected model of your data and operates on it:- Testing and monitoring. Anomaly detection, no-code cloud tests, and tests defined in code, run in your pipeline or via Elementary Runtime (optional).
- Column-level lineage across systems. From source through the warehouse to BI dashboards and AI consumption.
- Incident management. Related failures are grouped into one incident and routed by ownership, criticality, and impact, so one root problem does not turn into hundreds of alerts.
- AI agents and an MCP server. Backed by a context model of your whole stack, they handle triage, test recommendations, governance, and root cause analysis, and expose data health to your own AI tools.
Elementary does not require dbt. You connect your warehouse directly. If you do use dbt, the Elementary dbt package adds automatic collection of your dbt test and run results on top.
Editions
Elementary Cloud is the full platform. Elementary OSS is a free, self-hosted starting point for a single dbt project. You can begin with either and grow into more coverage as your needs evolve.Elementary Cloud
A fully managed platform for data observability, quality, and governance across your entire stack. It connects to your warehouse directly and covers many projects end to end. Use it to:- Integrate with your entire stack, from warehouse to BI, AI tools, and ticketing systems
- Give engineers, analysts, and stakeholders a shared view of what is trusted and what is broken
- Run checks in your pipeline as code, or via Elementary Runtime (optional)
- Reduce alert noise by grouping and routing alerts based on criticality, ownership, and impact
- Automate test recommendations, coverage, incident triage, and root cause analysis with AI agents
- Expose data health, lineage, and quality to your AI agents through the MCP server
- Enforce governance and support AI readiness, with enterprise security like RBAC, SSO, and audit logs
Elementary OSS
A free, open-source dbt package and CLI you host yourself, and a starting point for foundational observability on a single dbt project. It collects dbt test and run results, table-level lineage, and anomaly detection tests for freshness and volume, with basic alerts in a self-hosted report. Your data and metadata stay entirely in your environment. Set up Elementary OSS here. See the full OSS vs. Cloud comparison to understand the differences.Security and Privacy
Elementary is designed with security and privacy in mind.- Elementary Cloud does not have read access to raw data in your data warehouse.
- Elementary Cloud only extracts and stores metadata, logs and aggregated metrics.
- All data is encrypted at rest and in transit using industry standard protocols.
- Elementary uses service accounts or authentication tokens with granular and minimal permissions.
- Elementary offers deployment options with no direct access from Elementary Cloud to your data warehouse and third-party tools.
- Advanced authentication such as MFA, Okta SSO, Microsoft AD are available upon request.