SAN FRANCISCO, November 14, 2026 — Databricks has launched the second generation of Genie, an always-on autonomous agent designed to run complex enterprise tasks in the background and proactively bring decisions to humans. The upgrade shifts Genie from a reactive assistant to a system that delegates work to subagents, tracks progress across 100-plus data sources, and emails users when their input is required.
The Bottom Line
Genie now operates continuously on its own compute, managing long-running workflows such as quarterly forecasting or fraud monitoring without waiting for a prompt. A new priority inbox surfaces items needing approval, and email alerts keep users in the loop when they are away.
By the Numbers
Genie Search connects to more than 100 enterprise data sources, including the Databricks lakehouse, Google Drive, Jira, and Confluence. Thousands of customers already run Genie in production, among them Zillow, Unilever, adidas, Atlassian, and AstraZeneca. Databricks also disclosed plans to invest roughly US$3 billion across Europe and the Middle East over the next three years, its largest regional commitment outside North America.
How It Works
The agentic layer sits on Genie Ontology, a knowledge layer that became generally available this week after its June introduction. The ontology automatically maintains a company’s horizontal context — definitions, metrics, organizational structures, and processes — across structured and unstructured data regardless of where it lives. Unity Gateway gives administrators control over tool access, action tracing, and spending limits, keeping autonomy within existing governance boundaries.
Users assign a goal and set guardrails. Genie then takes ownership, spins up subagents in parallel when needed, follows up on missing inputs, and adapts to feedback. It can monitor insurance claims for fraud patterns, manage marketing campaigns around the clock, or prep a CEO for meetings by scanning email and Slack.
Analyst Take
“Taken together, I’d call [the new capabilities] significant because Genie stops waiting to be asked,” said Mike Leone, an analyst at Moor Insights & Strategy. “Handing it a goal instead puts answers in front of folks who would never open a dashboard in the first place.” Leone noted that while competitors including Snowflake, AWS, Google Cloud, and Microsoft are chasing similar autonomous-agent strategies, Databricks pulls ahead on the ontology layer because it learns business definitions from how people actually query data rather than requiring teams to write them out first.
What Comes Next
Genie and Genie Ontology are generally available now. The always-on agent, Genie Search, and Ontology Manager will roll out through preview and beta programs. Databricks co-founder and CEO Ali Ghodsi said the system has made him “10x more productive” by running in the background and returning only when a decision is needed. “It can do that because it knows how our company works: what we mean by revenue, which numbers are official, and who can see what,” Ghodsi said. “That’s the power of horizontal context, and Genie delivers it.”
Sources
- Databricks Launches New Always-on Genie: Autonomous AI for Complex Enterprise Work
- The new Genie: autonomous AI for complex enterprise work
- Databricks turns Genie into always-on agent for complex work
- Databricks upgrades Genie into an autonomous AI agent for enterprise workflows
- Databricks World Tour: Richard Shaw on Agentic AI Scale
Photo: Databricks Comms (CC BY-SA 4.0)
