Skip to content
Ready Mark

PLATFORM PROFILE, read 13 Nov 2026

Databricks Unity Catalog and AI-ready data: 7 of 10 conditions by rule, rarest check certified (7 of 17 tools); it trails OvalEdge most on the label reaching the agent (5 to 8)

See the full indexOpen the coverage matrix

Category
Lakehouse data catalog and governance layer (built into the Databricks platform, with an open source core)
Status
Exists under this name on 13 Nov 2026 (databricks.com/product/unity-catalog is titled "UNITY CATALOG Open and unified governance for data and AI"; docs.databricks.com pages last updated Nov 2026; open source repo unitycatalog/unitycatalog pushed 13 Nov 2026). Related surfaces have been renamed: Lakehouse Monitoring is now "data profiling" inside "Data quality monitoring", and agent traffic is governed through "Unity Gateway".
Warehouses named
Databricks SQL warehouses, Snowflake, BigQuery, AWS Glue, MySQL, PostgreSQL, Salesforce
Transformation named
dbt (dbt-databricks adapter), Lakeflow pipelines
BI named
Tableau, Power BI, Sigma, Databricks AI/BI dashboards
Rank and total
7 of 17, 58.0 of 100

7 of 10 checked by rule, 3 set by a person, 0 none found7 of 10 checked by rule, 3 set by a person, 0 none found

checked by rule set by a person none found on 13 Nov 2026

What Databricks Unity Catalog checks, condition by condition

Databricks Unity Catalog against the ten checklist conditions, from its documentation and public code. Read 13 Nov 2026.
ConditionHow it is checkedEvidence
Named ownerby a personDatabricks documents certifying a table by rule: owner is a named principal, the table has a description, and it had more than 100 read queries in 30 days, then the system.certification_status tag is set to certified. Source, read 13 Nov 2026.
Documentedby a personDatabricks documents certifying a table by rule: owner is a named principal, the table has a description, and it had more than 100 read queries in 30 days, then the system.certification_status tag is set to certified. Source, read 13 Nov 2026.
Certifiedby ruleDatabricks says tag automations (Beta) assign or remove governed tags on tables, columns and volumes from rules you define, and that runs update tags as data and metadata change. Source, read 13 Nov 2026.
Quality tests passby ruleDatabricks says anomaly detection monitors tables for freshness and completeness using per-table models built from commit history and row counts. Source, read 13 Nov 2026.
Freshby ruleDatabricks says anomaly detection monitors tables for freshness and completeness using per-table models built from commit history and row counts. Source, read 13 Nov 2026.
Sensitivity classifiedby ruleDatabricks says Data Classification automatically tags columns with class.* governed tags when automatic tagging is enabled, and tags all existing and future detections. Source, read 13 Nov 2026.
Lineage knownby ruleDatabricks says Unity Catalog captures runtime lineage to the column level for tables, jobs, notebooks, dashboards, models and external assets, with some limits. Source, read 13 Nov 2026.
Active useby ruleDatabricks documents certifying a table by rule: owner is a named principal, the table has a description, and it had more than 100 read queries in 30 days, then the system.certification_status tag is set to certified. Source, read 13 Nov 2026.
Metric definedby a personDatabricks says the Unity Catalog metric view lets users define a metric once and group by any available field at query time. Source, read 13 Nov 2026.
Access policy allows AI useby ruleDatabricks says ABAC policies are attached at catalog, schema or table level, evaluated dynamically, and support row filter and column mask policies, with GRANT and DENY policies (Beta). Source, read 13 Nov 2026.

Databricks Unity Catalog on the seven criteria

Blind scores for Databricks Unity Catalog, 0 to 10. A vendor claim scores at most 7 until a second source confirms it.
CriterionWeightScoreWhy
Labels by rule256Tag automations certify tables from owner, description and usage rules and re-run, but the feature is Beta and vendor-documented only.
Signals a rule can read207Rule conditions read tags, owner, description, upstream lineage tags, usage, freshness and quality results; vendor documentation only.
The label reaches the agent205Managed MCPs are governed with user identity and a PII guardrail (Beta); none found on 2026-11-13 for filtering by certification.
Reach across the stack155Tags cover tables, columns, views and AI/BI dashboards; BI tagging is Public Preview and Tableau or Power BI need manual registration.
Re-checks itself106Automations assign or remove tags on a schedule and classification refreshes within 24 hours; automation is Beta.
Evidence beyond the vendor57AWS Marketplace listing with ratings, Microsoft Learn pages and the Linux Foundation repo confirm the product; no third-party named customer or analyst page.
Price transparency54Pay-as-you-go per-second model published; no readable rate for governance features on 2026-11-13.

Where Databricks Unity Catalog is strongest

Rule-based labeling inside the platform: tag automations (Beta) can certify tables from owner, description and 30-day read-count rules, backed by automatic classification and anomaly detection.

Where Databricks Unity Catalog falls short

No documented way was found to filter or serve readiness tags to an agent over MCP, and tag automation and tag propagation are Beta.

Databricks Unity Catalog in each situation

Each situation ranks the same scores under its own published weights.
SituationRankTotalLeader
Board readiness7th, tied with Atlan, OpenMetadata and OvalEdge60.0Euno leads at 64.5
Metrics that disagree5th, tied with Secoda57.5Euno, OpenMetadata and OvalEdge tie for first at 60.5 (listed in that order by the published rule)
Agent this quarter656.0OvalEdge leads at 66.5
dbt, no owner list659.5OvalEdge leads at 63.5
Certified data only758.5OvalEdge leads at 66.5

The evidence on Databricks Unity Catalog, fact by fact

  • databricks-01 (Labels by rule, Databricks Unity Catalog's own claim): Databricks says tag automations (Beta) assign or remove governed tags on tables, columns and volumes from rules you define, and that runs update tags as data and metadata change. Source, read 13 Nov 2026.
  • databricks-02 (Labels by rule, Databricks Unity Catalog's own claim): Databricks documents certifying a table by rule: owner is a named principal, the table has a description, and it had more than 100 read queries in 30 days, then the system.certification_status tag is set to certified. Source, read 13 Nov 2026.
  • databricks-03 (Labels by rule, Databricks Unity Catalog's own claim): Databricks says system.certification_status marks an asset as certified or deprecated, with a check mark in Catalog Explorer, and that this can be automated at scale. Source, read 13 Nov 2026.
  • databricks-04 (Labels by rule, Databricks Unity Catalog's own claim): Databricks says Data Classification automatically tags columns with class.* governed tags when automatic tagging is enabled, and tags all existing and future detections. Source, read 13 Nov 2026.
  • databricks-05 (Labels by rule, Databricks Unity Catalog's own claim): Databricks lists the tag automation feature as Beta, with deterministic rules evaluated against metadata, and separate from the AI-based Data Classification. Source, read 13 Nov 2026.
  • databricks-06 (Signals a rule can read, Databricks Unity Catalog's own claim): Databricks documents tag automation conditions that read tags, descriptions, names, owner, dates, and the governed tags on direct upstream sources in lineage. Source, read 13 Nov 2026.
  • databricks-07 (Signals a rule can read, Databricks Unity Catalog's own claim): Databricks says the Insights tab shows table usage over 30 days and frequent queries, users, dashboards, notebooks and joined tables, and shows relative popularity of tables and columns. Source, read 13 Nov 2026.
  • databricks-08 (Signals a rule can read, Databricks Unity Catalog's own claim): Databricks says anomaly detection monitors tables for freshness and completeness using per-table models built from commit history and row counts. Source, read 13 Nov 2026.
  • databricks-09 (Signals a rule can read, Databricks Unity Catalog's own claim): Databricks stores data quality monitoring results (freshness, completeness, incidents with downstream impact) in a system table. Source, read 13 Nov 2026.
  • databricks-10 (Signals a rule can read, Databricks Unity Catalog's own claim): Databricks says Unity Catalog captures runtime lineage to the column level for tables, jobs, notebooks, dashboards, models and external assets, with some limits. Source, read 13 Nov 2026.
  • databricks-11 (The label reaches the agent, Databricks Unity Catalog's own claim): Databricks documents managed MCPs in system.ai (including a Genie One MCP and a Databricks SQL MCP) that are Unity Catalog securables invoked through Unity Gateway, with grants and policies. Source, read 13 Nov 2026.
  • databricks-12 (The label reaches the agent, Databricks Unity Catalog's own claim): Databricks says the Genie One MCP resolves metric definitions and table relationships through Genie Ontology, a governed semantic layer. Source, read 13 Nov 2026.
  • databricks-13 (The label reaches the agent, Databricks Unity Catalog's own claim): Databricks says the Genie Agent MCP bypasses Unity Gateway and relies on permissions on the Genie Agent and its data. Source, read 13 Nov 2026.
  • databricks-14 (The label reaches the agent, none-found): None found on 2026-11-13: a documented way for an MCP agent to filter or be served assets by system.certification_status or other readiness tags at query time, in Databricks docs for managed MCP servers, governed tags and tag automation Source, read 13 Nov 2026.
  • databricks-15 (Reach across the stack, Databricks Unity Catalog's own claim): Databricks says tags can be applied to tables, columns, volumes, views, functions, models, and also to dashboards, Genie Agents, Databricks apps and notebooks. Source, read 13 Nov 2026.
  • databricks-16 (Reach across the stack, Databricks Unity Catalog's own claim): Databricks says tagging of external metadata objects (the objects that stand for assets outside Databricks, such as BI dashboards) is in Public Preview. Source, read 13 Nov 2026.
  • databricks-17 (Reach across the stack, third-party): dbt Labs documents a databricks_tags config in dbt-databricks (version 1.11 or higher) that applies Unity Catalog tags to materialized views and streaming tables. Source, read 13 Nov 2026.
  • databricks-18 (Re-checks itself, Databricks Unity Catalog's own claim): Databricks says automations run on demand or on a recurring schedule, and a dry run records matched assets before any tag is assigned. Source, read 13 Nov 2026.
  • databricks-19 (Re-checks itself, Databricks Unity Catalog's own claim): Databricks says new tables and columns are typically classified within 24 hours, and a full scan can be triggered at any time. Source, read 13 Nov 2026.
  • databricks-20 (Evidence beyond the vendor, third-party): AWS Marketplace lists the Databricks Data Intelligence Platform as a SaaS product sold by Databricks, Inc., with a 4.6 rating from 1429 reviews shown on the listing. Source, read 13 Nov 2026.
  • databricks-21 (Evidence beyond the vendor, observed): The public GitHub repository for open source Unity Catalog describes the project as a sandbox project with LF AI and Data Foundation. Source, read 13 Nov 2026.
  • databricks-22 (Price transparency, Databricks Unity Catalog's own claim): Databricks says it uses a pay-as-you-go model billed per second with no up-front costs, and offers committed-use discounts. Source, read 13 Nov 2026.
  • databricks-23 (Price transparency, Databricks Unity Catalog's own claim): Databricks says Data Classification has its own billing and points readers to the pricing page; usage can be queried from system.billing.usage. Source, read 13 Nov 2026.
  • databricks-24 (Price transparency, third-party): The AWS Marketplace listing points to the Databricks pricing page for per-product DBU rates. Source, read 13 Nov 2026.
  • databricks-25 (Price transparency, none-found): None found on 2026-11-13: a DBU rate for Data Classification, tag automation, or Unity Catalog governance features as readable text, in the Databricks pricing pages (rate tables load by script and showed "Loading...") Source, read 13 Nov 2026.
  • databricks-26 (Profile, Databricks Unity Catalog's own claim): Databricks says external assets such as Tableau or Power BI dashboards can be registered as external metadata objects and linked into the lineage graph. Source, read 13 Nov 2026.
  • databricks-27 (Profile, Databricks Unity Catalog's own claim): Databricks says external metadata objects can have column names entered so column-level mapping to a Unity Catalog object is possible. Source, read 13 Nov 2026.
  • databricks-28 (Profile, Databricks Unity Catalog's own claim): Databricks shows Databricks AI/BI dashboards as downstream consumers on a table's Lineage tab. Source, read 13 Nov 2026.
  • databricks-29 (Profile, none-found): None found on 2026-11-13: automatic harvesting of Tableau or Power BI field and report lineage into Unity Catalog without manual registration, in Databricks external lineage and lineage documentation Source, read 13 Nov 2026.
  • databricks-30 (Profile, Databricks Unity Catalog's own claim): Databricks says row filters and column masks defined in Unity Catalog are enforced per user automatically in a Genie Agent, regardless of how it is shared. Source, read 13 Nov 2026.
  • databricks-31 (Profile, Databricks Unity Catalog's own claim): Databricks says MCP clients authenticate with a Databricks token that identifies a user or service principal, and OAuth user-to-machine calls the MCP as the signed-in user. Source, read 13 Nov 2026.
  • databricks-32 (Profile, Databricks Unity Catalog's own claim): Databricks documents a Sensitive Data Detection guardrail (Beta) on MCPs that can block calls containing PII in tool arguments and results, by classification tag. Source, read 13 Nov 2026.
  • databricks-33 (Profile, Databricks Unity Catalog's own claim): Databricks documents a custom policy function that denies a tool when the acting agent has a specific OAuth client ID and the call is on behalf of a user. Source, read 13 Nov 2026.
  • databricks-34 (Profile, Databricks Unity Catalog's own claim): Databricks logs AI/BI dashboard view events in the audit log, including a user accessing the published version of a dashboard. Source, read 13 Nov 2026.
  • databricks-35 (Profile, Databricks Unity Catalog's own claim): Databricks logs Genie One MCP tool calls in the audit log (one event per tool call). Source, read 13 Nov 2026.
  • databricks-36 (Profile, Databricks Unity Catalog's own claim): Databricks documents a query history system table with a column recording who ran each statement. Source, read 13 Nov 2026.
  • databricks-37 (Profile, none-found): None found on 2026-11-13: viewer logs for Tableau or Power BI reports built on Databricks tables, in Databricks system tables, audit log and lineage documentation Source, read 13 Nov 2026.
  • databricks-38 (Profile, Databricks Unity Catalog's own claim): Databricks says tag propagation (Beta) copies a governed tag from a source table or column to derived tables and views using lineage, when supported operations run on serverless compute. Source, read 13 Nov 2026.
  • databricks-39 (Profile, Databricks Unity Catalog's own claim): Databricks says propagation is off by default and enabled per governed tag, and uses a default value when sources conflict. Source, read 13 Nov 2026.
  • databricks-40 (Profile, Databricks Unity Catalog's own claim): Databricks says a propagated tag makes an existing matching ABAC policy, such as a column mask, apply to the derived column. Source, read 13 Nov 2026.
  • databricks-41 (Profile, Databricks Unity Catalog's own claim): Databricks says ABAC policies are attached at catalog, schema or table level, evaluated dynamically, and support row filter and column mask policies, with GRANT and DENY policies (Beta). Source, read 13 Nov 2026.
  • databricks-42 (Profile, Databricks Unity Catalog's own claim): Databricks says a Data Classification result can be turned into a masking policy: the form is pre-filled to mask columns with the reviewed classification tag. Source, read 13 Nov 2026.
  • databricks-43 (Profile, Databricks Unity Catalog's own claim): Databricks says Unity Catalog is open source under the Linux Foundation and federates Snowflake, BigQuery and AWS Glue without moving data. Source, read 13 Nov 2026.
  • databricks-44 (Profile, none-found): None found on 2026-11-13: a named integration with a third-party security tool (for example BigID) that feeds sensitivity labels into Unity Catalog tags, in the Databricks Data Classification, governed tags and ABAC documentation Source, read 13 Nov 2026.
  • databricks-45 (Profile, Databricks Unity Catalog's own claim): Databricks says the Unity Catalog metric view lets users define a metric once and group by any available field at query time. Source, read 13 Nov 2026.
  • databricks-46 (Profile, third-party): Microsoft Learn documents how to register Azure Databricks and interact with Azure Databricks Unity Catalog in Microsoft Purview. Source, read 13 Nov 2026.
  • databricks-47 (Profile, third-party): Microsoft Learn says a Purview scan of Unity Catalog reads Unity Catalog tags into Purview properties and fetches lineage between tables, views and columns during notebook runs. Source, read 13 Nov 2026.
  • databricks-48 (Reach across the stack, third-party): Tableau's own help documents a native Databricks connector for Tableau Cloud, Desktop, Prep and Server. Source, read 13 Nov 2026.

Sources

  1. docs.databricks.com/aws/en/admin/governed-tags/automate-tag-assignment, read 13 Nov 2026
  2. docs.databricks.com/aws/en/admin/governed-tags, read 13 Nov 2026
  3. docs.databricks.com/aws/en/data-governance/unity-catalog/data-classification, read 13 Nov 2026
  4. docs.databricks.com/aws/en/discover/table-insights, read 13 Nov 2026
  5. docs.databricks.com/aws/en/data-quality-monitoring, read 13 Nov 2026
  6. docs.databricks.com/aws/en/admin/system-tables, read 13 Nov 2026
  7. docs.databricks.com/aws/en/data-governance/unity-catalog/data-lineage, read 13 Nov 2026
  8. docs.databricks.com/aws/en/agents/mcp-tools/built-in-mcp-services, read 13 Nov 2026
  9. docs.databricks.com/aws/en/database-objects/tags, read 13 Nov 2026
  10. docs.getdbt.com/reference/resource-configs/databricks-configs, read 13 Nov 2026
  11. aws.amazon.com/marketplace/pp/prodview-wtyi5lgtce6n6, read 13 Nov 2026
  12. github.com/unitycatalog/unitycatalog, read 13 Nov 2026
  13. www.databricks.com/product/pricing, read 13 Nov 2026
  14. docs.databricks.com/aws/en/data-governance/unity-catalog/external-lineage, read 13 Nov 2026
  15. docs.databricks.com/aws/en/genie/set-up, read 13 Nov 2026
  16. docs.databricks.com/aws/en/agents/mcp-tools/authentication-networking, read 13 Nov 2026
  17. docs.databricks.com/aws/en/ai-gateway/govern-mcp-service, read 13 Nov 2026
  18. docs.databricks.com/aws/en/admin/account-settings/audit-logs, read 13 Nov 2026
  19. docs.databricks.com/aws/en/admin/system-tables/query-history, read 13 Nov 2026
  20. docs.databricks.com/aws/en/admin/governed-tags/tag-propagation, read 13 Nov 2026
  21. docs.databricks.com/aws/en/data-governance/unity-catalog/abac, read 13 Nov 2026
  22. www.databricks.com/product/unity-catalog, read 13 Nov 2026
  23. docs.databricks.com/aws/en/uc-semantics/metric-views, read 13 Nov 2026
  24. learn.microsoft.com/en-us/purview/register-scan-azure-databricks-unity-catalog, read 13 Nov 2026
  25. help.tableau.com/current/pro/desktop/en-us/examples_databricks.htm, read 13 Nov 2026

Last reviewed 13 Nov 2026.