Databricks Genie democratizes data access
Databricks
Databricks is turning the lakehouse from a tool used by specialists into a front end for the whole company. Genie shifts the key workflow from writing SQL, Python, and Spark jobs to asking questions in plain language, while still grounding answers in Unity Catalog data, permissions, and lineage. That matters because it lets Databricks sell more seats and more compute by moving from serving data teams to serving analysts, operators, and business users directly.
-
AI/BI Genie is built as a natural language layer on curated Unity Catalog data. Domain experts set up spaces with tables, sample queries, and instructions, then business users ask questions and get charts and SQL backed answers. The product removes query syntax as the bottleneck, but keeps governance and semantic setup as the control point.
-
This is also a competitive bundling move. dbt is pushing AI copilots so business users can do more of their own transformations, while ThoughtSpot built a search style analytics product for non technical users, and DataRobot and H2O.ai built no code model creation tools. Databricks is pulling those access layers into its own platform instead of leaving them to partners or point solutions.
-
The revenue implication is bigger than feature adoption. By late 2025, about 40% of Databricks revenue was already coming from expansion products like SQL and AI, not just core Spark and notebook workloads. Genie fits that same pattern, it makes the platform easier to enter, then monetizes through the governed compute, queries, apps, and downstream workflows users run after they ask the first question.
The next step is a fuller business user operating system on top of the data platform. Databricks One already packages dashboards, Genie, and apps into a simplified interface for non technical users, which points toward a future where the lakehouse is not just where data teams build, but where the rest of the company reads, asks, acts, and triggers AI workflows.