Databricks Announces Genie One, a New Agentic Partner for Every Department

Genie One supports business automation and coordination for departments such as marketing, finance, and sales, based on actual operational data.

The newly introduced Genie, powered by Genie Ontology, is a live context layer that continuously learns from internal and external data, AI tools, and key business applications, delivering more accurate answers faster and at lower token costs.

With the addition of Genie Agents and Genie App Builder, Genie’s overall capabilities are enhanced, enabling each department to build reusable agents and develop business applications using vibe-based coding, all under the governance of Unity Catalog.

Databricks (headquartered in San Francisco, California, USA; hereinafter “Databricks”), a company at the intersection of data and AI, today (June 16, US time) announced Genie One, a new agentic partner that supports automation and coordination across all data—structured, unstructured, analytical, and operational—both inside and outside the Databricks platform. Genie One is part of Databricks’ Genie family of AI partner products, transforming business data into trustworthy answers and actions.

At the core of the entire Genie ecosystem is Genie Ontology. This is a network that connects knowledge derived from every person and object within an enterprise—data, documents, tags, content, applications, records, and personnel information. One of the biggest challenges in enterprise AI is “correctly understanding the business from internal data and using AI more intelligently and effectively.” Genie Ontology is a new self-improving context layer that makes this possible. It automatically extracts and continuously updates business knowledge not only from data on Databricks but also from files, ticketing systems, chat, meetings, and other sources within connected AI tools and business applications. Based on this “ground truth” (authoritative reference data), Genie no longer guesses from incomplete context but retrieves actual answers from governed data and executes appropriate next actions. This results in improved answer accuracy, reduced latency, and lower costs.

Departments can now build reusable agents and applications connected to their own data, with built-in access control, permission management, and cost governance, by leveraging Genie Agents and Genie App Builder.

Why Early Enterprise AI Partners Failed to Meet Expectations

AI was able to revolutionize software development because the context developers need is consolidated in one place: source code. Source code is complete, structured, and easily readable by AI agents. However, similar transformation has not occurred in other enterprise departments because most business operations do not exist in such an environment. The required context is scattered across different systems, and much of the critical information exists only in people’s minds. Business departments want to ask simple questions about their situation and act based on the answers. But when information is fragmented and undocumented, even the most advanced AI agent cannot produce reliable responses. When context is missing, AI fills the gaps with guesses. In areas like finance, operations, and sales, confidently presented incorrect answers can be far more damaging than no answer at all.

Bridging the AI 'Context Gap'

Genie solves this problem by treating governed enterprise data—not documents or embedded data—as its primary information source. Genie Ontology continuously extracts and updates business context from across the Databricks platform and connected systems. As a result, Genie One does not infer from fragmented information scattered across multiple documents. Instead, it retrieves accurate answers to business questions from carefully selected, highly reliable data via SQL. This enables the creation of an agentic partner that deeply understands data and operates on the same data the company uses to run its business—explaining profit margin fluctuations, identifying upsell opportunities in the sales pipeline, and supporting financial closing processes.

Ali Ghodsi, Co-founder and CEO of Databricks, said:

“Much of today’s enterprise AI is simply guessing with misplaced confidence. That’s not good enough for business. If an AI can’t explain to a CFO why profit margins changed, or help a sales leader identify the next upsell opportunity, it’s not an AI problem—it’s a context problem. Genie Ontology continuously learns context from data everywhere, making responses much faster and agent accuracy much higher. This is the difference between a mere AI chatbot and an agentic partner that truly knows your business—its metrics, data sources, and answers. Every CEO and business leader should have access to Genie at all times.”

Detailed Information

Genie One is a data-savvy agentic partner available to everyone, accessible via web, iOS, and Android. While the original Genie offered only conversational analytics on data stored in Databricks, Genie One connects business users not only to data within Databricks but also to all data across a wide range of connected applications. It goes beyond answering questions to creating documents, reports, and deliverables. Genie One brings transformative change to data operations, both inside and outside Databricks. With an interactive interface featuring charts and graphs, it enables alert setup for continuous monitoring, task scheduling, creation of repeatable skills, and location-agnostic action execution via MCP tools.

The real-time, self-improving mechanism of Genie Ontology, grounded in enterprise context, forms the foundation for all of Genie’s reasoning. Genie Ontology automatically acquires and updates enterprise context from existing knowledge sources, building a true “ground truth” for business. This ensures that critical datasets and metrics are consistently understood, managed, and utilized. Genie integrates with major AI tools and over 50 common business applications and data systems,