Published: November 13, 2025
by Peter Wang, Hongyi Zhang, Anthony Hong, Tao Feng and Kelly Albano
Every analyst knows the feeling: you open a dataset in your catalog, scroll through columns, and wonder: “Is this even the data I need?”. Until now, answering that question meant writing exploratory SQL, checking lineage, or tracking down documentation.
With the new Sample Data Exploration experience in Unity Catalog, powered by Databricks Assistant, you can simply ask your data. Just type a question in natural language, like “Which region has the highest sales?”, and get instant answers or visualizations, right on the sample data page within the Catalog Explorer UI.
This capability is now in Public Preview, marking another step forward in the Databricks Data Intelligence Platform, where AI and context work together to help every user, whether technical or not, move from data to decision faster.
Most data platforms still rely heavily on technical experts to interpret the meaning of data, its origin, and how to utilize it. Without built-in intelligence, teams waste time searching for the right datasets, validating trust, and translating technical structures into business concepts.
Unity Catalog changes that by serving as the intelligent foundation of the Databricks Data Intelligence Platform, connecting business context with underlying data, models, and lineage.
The new Sample Data Exploration capability brings that intelligence directly into the Catalog Explorer experience. On any dataset’s Sample Data tab, you can now:

The Databricks Assistant isn’t a generic chatbot; it’s a context-aware copilot that uses metadata, lineage, and governance signals from Unity Catalog to ensure every response is grounded in your organization’s trusted, governed data.
This feature bridges the gap between finding data and understanding it, helping every user quickly assess relevance, accuracy, and value.
Sample Data Exploration is part of a broader set of AI capabilities that make Unity Catalog more than a governance layer, but a Data Intelligence Engine that continuously enriches, protects, and optimizes your data. Recent innovations include:
AI-Generated Comments: Automatically create descriptive table and column comments to improve dataset context. Recently migrated to the Assistant platform for higher quality and unified model control, this update has driven a 36% increase in accepted or edited comments and saves roughly $200K annually in internal model serving costs.
Bulk Column Comments: A new modal allows users to review and apply AI-generated comments across all columns in a table, resulting in a 6× increase in weekly throughput and a 400% increase in tables with at least one AI-generated comment.

Databricks Data Classification: Detect and protect sensitive data automatically with policy-driven access controls. Learn more in our announcement blog!
Unity Catalog Managed Tables: Simplify performance and cost management with an intelligent storage layer that delivers up to 20× % faster queries and 50% lower costs while unifying governance and observability in one place. Our blog provides a detailed breakdown of these benefits.
Unity Catalog Business Semantics: Bring a unified, governed, and open semantic foundation to power consistent and trusted insights across all your BI assets, developer tools, and AI agents. Learn more here.
Together, these capabilities make Unity Catalog the intelligent foundation of the Databricks Data Intelligence Platform, one that not only governs your data but continuously learns from it.
Getting started is simple:
If you’re ready to explore how AI and governance come together in Unity Catalog, you can also:
