Databricks has announced an always-on version of Genie that it says can manage long-running enterprise tasks, monitor changes and ask people for decisions through a priority inbox and email. The November 11 announcement expands Genie beyond answering questions, although the new always-on capabilities are not generally available yet.
According to the company’s official announcement, the new always-on Genie, Genie Search and Ontology Manager are coming to preview and beta programmes. Existing Genie and Genie Ontology are generally available to Databricks customers.
From answering questions to tracking work
Databricks says employees will be able to give Genie a goal and boundaries, then let it track progress, follow up on missing information and return when a decision is needed. Examples include monitoring changes in customer usage, preparing meeting briefs and tracking revenue trends.
A priority inbox brings items needing attention into one place. Email alerts link back to Genie, with updates on the same issue kept in one thread.
Business context and access controls
Genie Ontology provides the business context behind the work, such as agreed definitions of revenue, relationships between data and the people involved in a process. Genie Search is intended to bring together information from Databricks data and connected applications such as Google Drive, Jira and Confluence.
Databricks says Unity Catalog applies data-access permissions, while Unity Gateway lets administrators control tools, approval rules and spending limits and trace actions. Those controls matter when an AI system can act across business applications, rather than just produce an answer.
The company also describes support for creating data-driven slides, dashboards and reports, and working through collaboration tools. It has not given a firm release date for the new always-on preview in the announcement.
Availability and cost
Genie is part of the Databricks Data + AI Platform. Databricks describes pricing as usage-based rather than per seat, but the announcement does not provide a fixed price for the new capabilities. Businesses evaluating them should check preview access, connector permissions and usage costs before treating the launch as a ready-to-deploy autonomous service.
