Skip to main content

Data governance

At a glance

Governance centralizes control, quality, and traceability of your data. Define approval policies, classify sensitive information, monitor instance quality, and visualize data lineage — all from a unified interface.

Practical example: in a procurement process, you can require dual approval before any modification to a supplier contract, classify prices as "Confidential", verify that every order has a positive amount, and trace the origin of each piece of data in the catalog.

What you'll do

  • Define approval policies with configurable quorum (one or more validations required)
  • Classify your data by sensitivity level (Public, Internal, Confidential, Highly Confidential)
  • Create automated quality checks (non-null, uniqueness, range, regex, freshness)
  • Visualize data lineage and manage your business catalog
  • Review operation logs to audit any modification
  • Manage member roles and permissions with custom roles and individual overrides
  • Control access to specific resources (spaces, sources, documents) via ACLs

Key vocabulary

TermMeaningExample
PolicyA rule that defines when an approval is required."Any instance deletion requires 2 approvals"
ClassificationA sensitivity level assigned to a piece of data.Confidential, Public, Internal
Quality checkAn automated verification on your instances."The email field must match a valid format"
LineageThe trace of a data item's origin and journey.Supplier -> Order -> Invoice
CatalogThe inventory of your entity types with their metadata.Owner, steward, quality score, freshness
QuarantineAn isolation zone for instances that fail a check.Instance "Order #42" awaiting correction
Custom roleA role you create with specific permissions."Analyst": read dashboards + knowledge base
OverrideA permission adjustment for a specific member.Deny ontology access for Marie
ACLAccess control list on an individual resource.Space "Finance": read-only for Jean
  • Ontologie: governance applies to the entity types and instances of your business model. Approval policies protect critical modifications.
  • Workflows: pipelines can automatically trigger approval requests when a governed operation is detected.
  • AI Agent: the agent respects governance policies and can query the data catalog to answer questions about data quality or origin.

Expected result

Your data is protected by approval policies, classified by sensitivity level, monitored for quality, and traced in a centralized catalog. Every critical modification goes through a validation workflow before being applied.

Need help?

Contact us: Support and contact.