Databricks has introduced an upgraded version of its Genie AI agent, enabling it to perform complex, multistep business tasks with minimal human supervision. Announced on November 11, 2026, the update uses a new ontology system to help the agent understand and analyze information stored across a company’s Databricks platform.
The company is positioning Genie as an AI assistant for executives who need to evaluate business performance, identify operational problems, and make decisions using internal company data. Databricks CEO Ali Ghodsi demonstrated how he uses the technology to analyze employee feedback, review unsuccessful sales opportunities, and prepare for interviews.
Genie Uses Business Context to Complete Multistep Tasks
The updated Genie uses an ontology developed by Databricks to understand relationships between information stored in a company’s systems. This provides additional context that allows the AI agent to complete tasks involving multiple datasets and extended analysis.
Rather than responding only to individual questions, Genie can carry out workflows that require several steps to reach a result. Databricks says the agent can perform these tasks independently after receiving instructions from users.
The technology is designed to help business leaders obtain insights from information that may otherwise require coordination between departments or manual analysis. Its ability to work across different business records also allows executives to examine issues from multiple perspectives.
Databricks CEO Demonstrates Internal Applications
Ghodsi demonstrated Genie’s capabilities using his own business conversations and reports rather than a prepared product demonstration. One example involved analyzing employee satisfaction survey results to identify areas where Databricks received lower scores.
The agent also helped evaluate proposed solutions from company executives, identifying suggestions that appeared less likely to address the underlying problems. In another example, Genie summarized an unsuccessful sales presentation and identified product capabilities that a prospective customer wanted before making a purchase.
Ghodsi also used the agent to prepare for a podcast appearance by reviewing previous interviews and generating potential questions. During a separate conversation, he asked Genie for suggestions on how to respond to the interviewer.
AI Agents Could Help Executives Identify Internal Problems
Ghodsi said AI agents could become particularly useful as organizations grow beyond the point where executives can maintain direct relationships with employees. He referenced Dunbar’s number, a theory suggesting that people can maintain approximately 150 stable social relationships.
With around 12,000 employees at Databricks, Ghodsi acknowledged that he cannot personally communicate with everyone in the organization. He sees Genie as a way to access information about unsuccessful projects, employee concerns, and operational issues that might otherwise be difficult for senior leadership to identify.
The company is also addressing concerns surrounding the operating costs of continuously running AI agents. Ghodsi said Databricks’ Unity Gateway allows organizations to establish usage limits and select less expensive open-weight AI models when appropriate.
Databricks expects the updated Genie to support business workflows in ways similar to how AI coding agents assist software developers. The company has not disclosed specific performance benchmarks for the updated system.
Featured image credits: Wikimedia Commons
For more stories like it, click the +Follow button at the top of this page to follow us.