Built for data engineers who care about correctness
Decube lives in your existing workflow. No new dashboard to train your team on. Just context, surfaced where you already work.
Where data teams use Decube daily
Pipeline incident response
When a Slack alert fires at 9am saying the revenue table has 40% fewer rows than yesterday, Decube shows the upstream source that stopped loading at 2am - in under 3 minutes.
Safe schema changes
Before renaming a column in orders.customer_id, run an impact analysis in Decube. It shows the 14 dbt models and 3 Tableau dashboards that reference it, and who owns each one.
Data stewardship
Data stewards certify business definitions in Decube's glossary. Finance and data teams look at the same definition of "ARR" - linked to the exact revenue calculation logic.
Data quality coverage
New table lands from a Fivetran connector. Decube auto-suggests quality rules based on the column data types and historical patterns. One-click activation. No SQL required.
Compliance and audit
SOC 2 and GDPR audits require proof of data lineage. Decube generates column-level provenance reports on demand - showing the path from raw source to any report field.
Team onboarding
New data engineers spend their first week asking "what does this table do?" Decube's search surfaces column descriptions, ownership, and quality history - cutting onboarding from two weeks to two days.
Fast search. Deep context. Zero noise.
Type any table name, column, or business term. Decube returns the definition, quality status, recent changes, lineage upstream and downstream, and the team who owns it - in one panel.
The lineage graph is interactive. Click any node to see its quality metrics. Click any column to see every transformation that produced it. It is the first lineage tool your analysts will actually use.
Book a Demo
Fits your stack without changing it
Read-only access
Decube never writes to your warehouse. It connects through read-only service accounts and queries information schema and query history views. Your data stays where it is.
SaaS or self-hosted
The managed SaaS version runs in AWS us-east-1 with SOC 2 Type II certification. The self-hosted option deploys to your VPC via Helm chart - same features, your infrastructure.
REST API and Python SDK
Trigger lineage scans from your dbt CI pipeline. Push custom metadata via the REST API. The Python SDK supports PyPI installation and integrates with standard data engineering toolchains.
SSO and RBAC
Okta, Azure AD, and Google Workspace SSO supported on all paid plans. Role-based access control lets you restrict sensitive table metadata to specific teams.
See Decube with your own data
We connect to your warehouse during the demo. You will see your actual lineage graph in 10 minutes.
Book a Live Demo