Although Databricks never uses the term CoPilot, the company says it has launched Genie One Agent Coworker as a member of its Genie suite of coworking agents that turn business data into answers and action.
The suite is based on the Genie ontology and provides data context, a web of all knowledge from everyone and everything in your organization, including data, documents, tags, content, apps, documents, and people. According to Databricks, this is a self-improving context layer that helps you fully understand your business from your data. Customer AI developers can use Genie Agent and Genie App Builder to create reusable agents and applications. All of these are connected to data with built-in access controls, permissions, and cost governance.
“Most enterprise AI today just makes guesses with false confidence, which is not good enough for business,” said Ali Ghodsi, co-founder and CEO of Databricks. If you can’t find your next upsell, it’s not an AI issue, it’s a context issue. Genie Ontology continuously learns context from data everywhere, so answers are much faster and agents are more accurate. That’s the difference between AI chatbots And agent colleagues who know your business inside and out – they know every metric, every data source, every answer. Every CEO and business leader should have Genie at their fingertips. ”
Genie Ontology automatically extracts and continuously updates business knowledge from Databricks as well as AI tools for business users across databases and over 50 popular apps and data systems. Files, tickets, chats, meetings. Build a knowledge graph of your data. Current integrations include Google Drive, Jira, Slack, Confluence, SharePoint, and more. It is our understanding that enterprise data assets are ranked using the OntoRank concept. Consider Google PageRank for business knowledge entities in an ontology.
This new ontology context layer enables Genie to provide more accurate answers while reducing latency and reducing data-intensive costs. question. This provides the ground truth, a master data record so that different AI agents working on an organization’s corpus of data share common definitions and background information.
The first generation of Genie provided conversational analytics on data stored in Databricks. Genie One goes even further, connecting users to Databricks data A wide range of connected applications. This is a step change in automating data work beyond just answering questions and creating documents, reports, and deliverables. With visual interfaces like interactive charts and graphs, users can use MCP tools to set constant monitoring alerts, schedule tasks, create repeatable skills, and take actions from anywhere.
detail:
-
Genie One is available on the web, iOS, and Android.
-
Genie agents allow teams to save Genie conversations as reusable agents that inherit the conversation’s memory, including its source, instructions, and behavior, allowing colleagues to recall the conversation by name and repeat workflows across teams. Employees can create agent skills and share them with each other for consistent and repeatable response formats and workflows.
-
Genie App Builder is a fully managed vibe coding environment. Teams can upload their business context and Genie App Builder will generate a live build plan and preview of a working app Connect to real, managed enterprise data. These applications (built for internal teams or customers) run securely and cost-effectively with Unity Catalog permissions and access controls built in from the beginning.
-
Genie Code is an autonomous AI agent that helps data teams plan, build, and execute data engineering, machine learning, and analytics workflows. that Includes a dedicated workspace to track your progress. Review the steps and switch threads to work between projects.
-
Genie ZeroOps is a new background agent built into Databricks that autonomously monitors, inspects, and suggests fixes for your data and AI assets, including pipelines, jobs, tables, and ML models.
For more information about Genie, please visit Databrick Blog.
open sharing
Databricks also launched OpenSharing, an open standard for sharing data and AI assets across organizations and platforms, hosted by the Linux Foundation. It is built on Delta Sharing, a zero-copy way for Databricks to use its own data to share third-party data with customers.
OpenSharing says it introduces the first open, vendor-neutral protocol for securely sharing AI assets such as agent skills, AI models, and unstructured data, enabling secure collaboration and monetization. Assets in the AI era. By adding support for the Iceberg IRC (Iceberg REST Catalog) client, we extend the extensive cross-platform delta sharing ecosystem and extend the reach of data providers to more recipients. Customers can now access on-premises or private cloud assets via OpenSharing through storage partners such as Everpure, MinIO, and Qumulo.
availability
Genie has no seat-based pricing and organizations only pay for the AI they actually use, with up to $10 free per user per month. Genie One, Genie Agents, and Genie Code are now generally available. genie app builder And Genie ZeroOps will go into private preview immediately after the Data + AI Summit. Genie is now available as a native app for iOS and Android.
boots notes
Ontology is the philosophical study of existence, which has a structure that separates it into unique particulars and common universals, as well as into concrete and abstract entities.
