GUIDE, STACK, read 13 Nov 2026
Databricks can check all ten AI readiness conditions natively, and seven of them by rule
Databricks has a native feature for every one of the ten conditions
This guide reads Databricks' documentation on 13 Nov 2026 and counts what Unity Catalog can check without another product. Seven conditions can be checked by a rule Databricks runs (certified, quality, fresh, sensitivity, lineage, usage, access). Three rest on a person (owner, documented, metric definition). None of the ten returned none found on 13 Nov 2026, so the count has no gap column to fill.
The caution is maturity. Rule-based certification depends on tag automations, which Databricks marks Beta, and the limits are named below. On this site Databricks ranks 7th of 17 with 58.0 points.
Governed tags carry certification, and a rule can set it
Per Databricks' tags documentation, "Governed tags are account-level tags with enforced rules for consistency and control." You define the allowed keys and values and decide who may assign them. The certification tag is system.certification_status.
The tag automation page goes further than a manual tag. It lists "Certifying trusted data that meets your readiness criteria" as a use, and its worked example matches a table when "The owner is one of the principals responsible for production data. The table has a description. The table had more than 100 read queries in the last 30 days." The action assigns system.certification_status with the value certified. That is a readiness rule written as data. A dry run records matched assets before anything is tagged, so you can read the result before it changes a label.
Comments and ownership are plain Unity Catalog properties that the rule reads, so a table missing either never qualifies. Check them with a query on the catalog's information schema before you rely on the rule.
Anomaly detection and system tables cover quality, freshness and usage
Databricks' data quality monitoring page states: "Anomaly detection monitors enabled tables for freshness and completeness." It builds a per-table model from the history of commits and flags a commit that is unusually late.
For usage, the Insights tab shows a table's use over 30 days, and the documentation adds a limit: "The queries listed on the Insights tab are limited to queries performed using Databricks SQL." The query history system table has an executed_by column, described as "The email address or username of the user who ran the statement." Viewer logs for Tableau or Power BI reports built on Databricks tables returned none found on 13 Nov 2026.
Classification, ABAC and metric views complete the set
Sensitivity: the Data Classification page says it "uses an agent to automatically classify and tag tables in your catalog." The page says it uses a large language model, that usage is queryable from system.billing.usage, and that a catalog owner must enable it.
Access: attribute-based access control, per Databricks, means "access is determined by evaluating attributes associated with securable objects." Tags drive the policy, so a classification tag can lead directly to a mask.
Metric definition: Databricks says metric views "provide a centralized way to define and manage business metrics." A person writes the definition, so the check is that the certified table has a metric view and the agent's metric is in it. Lineage is native and covered on another site.
The ten conditions, checked with Databricks alone
The table maps each condition to the Unity Catalog feature found in the documentation.
| Condition | What the platform gives you | Gap |
|---|---|---|
| Named owner | Owner property on every Unity Catalog object, readable by tag automation rules. | Set by a person; a team alias still passes. |
| Documented | Comments on tables and columns, readable by rules. | Set by a person. |
| Certified | system.certification_status, set by hand or by tag automation. | Automation is Beta. |
| Quality | Anomaly detection for completeness. | Covers enabled tables only. |
| Fresh | Anomaly detection learns each table's commit pattern. | Per-table model, not a stated deadline. |
| Sensitivity | Data Classification writes class.* governed tags. | LLM-assisted; cost in system.billing.usage. |
| Lineage | Runtime lineage to column level (covered elsewhere). | External BI assets are registered by hand. |
| Usage | Insights tab, popularity, query history table. | Insights covers Databricks SQL queries only. |
| Metric definition | Unity Catalog metric views. | Set by a person. |
| Access | ABAC policies, row filters and column masks driven by tags. | Governed tags needed first. |
Databricks leaves outside assets and the agent-side check to another tool
Two gaps stand out. Labels stop at Unity Catalog: automatic harvesting of Tableau or Power BI fields into Unity Catalog without manual registration returned none found on 13 Nov 2026, and external BI assets are a manual registration step. Delivery to the agent also stops short: a documented way for an MCP agent to be served assets by system.certification_status at query time returned none found on 13 Nov 2026.
A team that runs Snowflake or BigQuery beside Databricks needs one place to compare labels. The ranking below lists the tools whose evidence names Databricks.
17 of 17 ranked tools name Databricks in their evidence
This table is built from the systems each tool names, read 13 Nov 2026. A tool that names Databricks can label its assets only as far as its own documentation shows; the marks come from the Readiness Coverage Matrix.
| Rank | Tool | Named as | Reach | Conditions by rule |
|---|---|---|---|---|
| 1 | OvalEdge | Databricks | 5 | 6 of 10 |
| 2 | Euno | Databricks | 6 | 8 of 10 |
| 3 | DataHub | Databricks | 5 | 6 of 10 |
| 4 | OpenMetadata | Databricks | 6 | 7 of 10 |
| 5 | Atlan | Databricks | 6 | 8 of 10 |
| 6 | Secoda | Databricks | 5 | 8 of 10 |
| 7 | Databricks Unity Catalog | Databricks SQL warehouses | 5 | 7 of 10 |
| 8 | dbt | Databricks | 5 | 6 of 10 |
| 9 | Alation | Databricks | 5 | 7 of 10 |
| 10 | Ataccama | Databricks | 4 | 5 of 10 |
| 11 | Qlik Talend Cloud | Databricks | 3 | 6 of 10 |
| 12 | Monte Carlo | Databricks | 4 | 5 of 10 |
| 13 | Snowflake Horizon | Databricks | 3 | 6 of 10 |
| 14 | Precisely | Databricks | 3 | 5 of 10 |
| 15 | Actian | Databricks Unity Catalog | 2 | 6 of 10 |
| 16 | Informatica | Databricks (Unity Catalog) | 2 | 7 of 10 |
| 17 | Collibra | Databricks Unity Catalog | 3 | 5 of 10 |
Sources
- Databricks: Automate tag assignment, read 13 Nov 2026
- Databricks: Apply tags to Unity Catalog securable objects, read 13 Nov 2026
- Databricks: Data quality monitoring, read 13 Nov 2026
- Databricks: Data Classification, read 13 Nov 2026
- Databricks: Table insights and popularity, read 13 Nov 2026
- Databricks: Unity Catalog metric views, read 13 Nov 2026
- Databricks: Query history system table, read 13 Nov 2026
- Databricks: Attribute-based access control, read 13 Nov 2026
Last reviewed 13 Nov 2026.