Databricks has introduced Genie One, a new enterprise agent platform designed to help business users automate work and act on company data. The product was announced at the company’s Data + AI Summit and is positioned as part of a broader Genie suite aimed at turning enterprise data into answers, actions, and applications.
The company says Genie One is meant for teams in areas such as marketing, finance, sales, and operations. Rather than focusing only on analytics, the platform is intended to help users orchestrate tasks across structured and unstructured data, whether that information sits inside Databricks or in connected business systems.
At the center of the release is Genie Ontology, which Databricks describes as a live context layer that continuously learns from an organization’s internal and external data, along with AI tools and workplace apps. The company says the system is designed to solve a major challenge in enterprise AI: giving models enough reliable context to answer questions accurately.
Databricks said Genie Ontology draws from sources including files, tickets, chats, meetings, and other workplace systems. The goal is to create a continuously updated view of business knowledge so Genie can rely on governed data rather than trying to infer answers from incomplete information. Databricks says this approach should improve accuracy, reduce latency, and lower token costs.
The company has also added integrations with more than 50 popular apps and data systems. Examples named in the announcement include Google Drive, Jira, Slack, Confluence, and SharePoint.
Alongside Genie One, Databricks unveiled Genie Agents and Genie App Builder. Genie Agents lets users save a conversation as a reusable agent, including its memory, instructions, and sources. That means teams can call on a trusted workflow by name and reuse it across the organization.
Genie App Builder is aimed at enterprise app creation. Databricks says it provides a managed environment where teams can upload business context and generate a live build plan and working app preview connected to governed data. The company says the applications will be protected by Unity Catalog permissions and access controls.
Databricks also introduced Genie Code, an autonomous agent for data teams working on engineering, machine learning, and analytics tasks. A separate feature, Genie ZeroOps, is designed to monitor data and AI assets such as pipelines, jobs, tables, and models, then propose fixes when it finds problems.
The launch reflects Databricks’ effort to move beyond conversational analytics toward a wider enterprise AI platform. The company argues that business work is harder for AI to support than software engineering because crucial context is spread across many systems and often remains undocumented.
Databricks chief executive Ali Ghodsi said many enterprise AI systems fail because they guess when context is missing. He said the new context layer is intended to make responses faster and agents more accurate.
Several customers were cited in the announcement as early users of the platform. Albertsons Companies said it is using Databricks to support merchandising intelligence. Uplight said Genie One is helping teams explore and innovate with more speed and confidence. Foot Locker said Genie Agents are supporting executives and business teams with centralized AI-driven insights.
Genie One, Genie Agents, and Genie Code are generally available now, according to Databricks. Genie App Builder and Genie ZeroOps are slated to enter private preview after the summit. The company said Genie is available on the web as well as native iOS and Android apps, and that it does not use seat-based pricing.