We’re excited to introduce Agent Mode to Genie Space. Our team has developed a powerful agentic process that iteratively plans, explores, and reasons over your data to answer your business questions. As part of Databricks’ Week of Agents, this blog highlights another way we are changing how organizations interact with their data using agentic AI.
This experience opens up a more practical level of data analysis to everyone in your organization. Now, anyone can get real-time insight into complex business questions, like:
- Why did our churn rate increase in the third quarter?
- How can we optimize our campaign spend?
- What revenue impact should I expect if these two supply lines are disrupted?
How does agent mode work?
When you ask a question in agent mode, the Genie doesn’t just return a question. It examines the problem like a real data analyst: planning an approach, testing hypotheses, and iterating toward an explanation.
For example, imagine a Genie space for customer support. You look at the increase in reopened cases in December 2025 and ask: “What is contributing to the increase in reopened support cases?”
Agent mode first confirms the spike, then searches for potential contributors such as customers, products, categories or teams. It uses the business context in your Genie space, including Unity catalog metadata and author-defined semantics, to focus on the most relevant potential contributing factors.
The agent mode evaluates these hypotheses by executing multiple queries against the underlying data. Semantics defined in Jini repository of knowledge Teach it to generate accurate questions. We have also made the working of Agent Mode transparent so that users can always verify its accuracy.
During its analysis, the Genie continuously considers the results of each query and decides what to explore next. In this example, after examining several potential drivers of the spike, the agent decides that it should further investigate whether seasonal patterns are contributing to the spike. This iterative cycle of hypothesis generation, questioning, and reflection allows Genie to explore the data more deeply and arrive at a well-supported explanation.
After completing its analysis, the Genie generates a report of its findings. Following this example, the report first quantifies the increase in reopened cases and then identifies the primary contributors – namely the increase in bug-related cases and the performance of the L2 regional team. To support these findings, the report also includes visualizations and references to the underlying SQL for users to review.
Depending on the type of question, Genie also provides actionable recommendations on what teams should focus on to improve performance. Users can share these reports directly on the platform or download them as PDFs to easily distribute and collaborate on insights.
Built for questions of any complexity
Agent Mode not only unlocks advanced business checks – it improves accuracy across all types of queries, from simple analysis to multi-step analysis.
Even for straightforward questions, he takes small validation steps to make sure he understands the data before answering. We designed the agent to dynamically scale its reasoning according to the complexity of the task – moving faster for simple prompts, and spending more time planning and evaluating for deeper investigations.
The result is faster answers to everyday questions and more rigorous analysis when dealing with complex problems.
Users interacting with Genie inside an AI/BI dashboard can also take advantage of this new experience. When you ask questions to Genie from the dashboard, it leverages agent mode by default.
Get started today
Workstation administrators can now confirm that Agent Mode is enabled workspace preview Page. Once enabled, simply turn on the Agent toggle in your Genie space and ask your business questions.
Agent mode is available now, with more to come, including API support and unstructured document analysis. Try Agent Mode today – we can’t wait to hear what you think.
