Find the table, then find out whether to trust it.
Discovery is the easy half. The catalog also carries where a table came from, who owns it, what it costs to produce and who is allowed to read it.
Lineage from the engine
OpenLineage events come out of Apache Spark™ and Trino™ as they run. Nothing is inferred from parsing code after the fact.
Domains with their own admins
A domain separates namespaces and has its own admin and team. Finance does not need a platform ticket to grant access to a finance table.
Marketplace for discovery
Publish a dataset, describe it once, let other teams request access against it. Shared compute can be granted the same way.
Sharing outside the account
Publish a table into every catalog's shared namespace and grant a scoped Polaris role against it. Partners and cross-account teams read over the Iceberg REST catalog with vended, table-scoped credentials, not a copied dataset.
PII found before it's asked about
Profiling scans column shapes for PII patterns and tags the result against the W3C DPV vocabulary, reviewed by a person before the tag affects anything. See Security for how tags turn into masking.
Learn more →See it against your own data
We will walk through your workloads and show what runs where.