OpenAI's Data Agent Turns Questions Into Power BI and Tableau Dashboards
The new Data agent in ChatGPT Work can connect business data, investigate changes, and build interactive dashboards from plain-language prompts.
OpenAI Data Agent is now available in ChatGPT Work, letting users turn plain-language questions into data analysis and interactive dashboards. The agent can connect to approved company data, investigate what changed, explain the evidence behind its findings, and build dashboards without requiring users to write queries or learn a separate analytics workflow.
The release was announced by OpenAI on September 10, and the company says the Data agent is available through the Plugin directory in ChatGPT Work. Workspace administrators can install it for teams and control which connected data-source plugins are available.
The Data agent connects to business data and context
The Data agent can connect to sources including Amazon Redshift, Datadog, Google BigQuery, ClickHouse, Databricks, MongoDB, Snowflake, and other approved systems. It can also bring files and documents from Google Drive and SharePoint into an analysis.
OpenAI says the agent can use an organization's business terms, metric definitions, custom calculations, and data relationships. Those definitions can come from semantic layers and trusted sources such as Databricks Genie Ontology, dbt, GitHub, Snowflake Horizon, and existing BI dashboards.
- Connect approved business data and supporting documents.
- Investigate changes and ask follow-up questions in the same conversation.
- Review the evidence behind findings before acting on them.
- Turn analysis into interactive dashboards that teams can edit, share, and refresh.
Power BI and Tableau dashboards can be built from plain language
The most visible part of the launch is the ability to move from a question to a dashboard without manually building every report. OpenAI says the Data agent can build and interact with dashboards in Omni, Oracle BI, Power BI, Sigma, Tableau, and ThoughtSpot.
That does not mean the agent replaces every existing business-intelligence system. Instead, OpenAI is putting a conversational layer in front of tools teams already use. A user can ask a follow-up question, refine the analysis, and direct the work in natural language while keeping the resulting dashboard inside an existing BI workflow.
The company also says teams can share brand guidelines with the Data agent so generated dashboards can be tailored to an organization's visual style.
OpenAI is positioning Data for business teams, not only analysts
OpenAI says nearly all of its product team and more than two-thirds of its go-to-market organization use data agents in ChatGPT Work to analyze company data. The company says its internal deployment relies on shared business definitions, access rules, and safeguards for sensitive data.
Organizations in OpenAI's Alpha program, including NTT DATA, Thermo Fisher, ServicePiston, and others, are also using the Data agent for tasks such as analyzing sales and spending, catching reporting errors, and deciding which opportunities to pursue.
The launch fits into OpenAI's broader expansion of ChatGPT from a question-and-answer interface into a system that can complete longer work across connected apps and files. Saganote recently covered OpenAI's ChatGPT Work and Codex admin controls, which are relevant to how organizations manage these connected capabilities.
Data analysis joins a wider push inside ChatGPT
The Data agent is arriving alongside a series of new ChatGPT capabilities aimed at specialized work. OpenAI has also been expanding access to information and tools across areas such as healthcare, where Saganote recently covered ChatGPT for Healthcare, Epic EHR, and public data.
The company's consumer product is changing at the same time. Recent updates to ChatGPT Images 2.5 added faster image generation and more precise editing, showing how OpenAI is adding more task-specific capabilities alongside the core chat experience.
OpenAI is also expanding the commercial side of ChatGPT. Saganote's coverage of ChatGPT ads, the $1 billion revenue milestone, and global expansion provides context for why business-oriented features such as Data are becoming an increasingly important part of the product.
The Data agent is not simply an open web analytics tool. Workspace administrators control its installation, while connected data sources can require separate setup or authorization. The agent uses the permissions of the connected accounts.
How teams can get started
- Install Data from the Plugin directory in ChatGPT Work, if it is not already available.
- Have the workspace administrator enable the required data-source plugins and apps.
- Complete any required account-connection or authorization steps.
- Start a conversation with
@Dataand ask a business question. - Review the analysis and evidence, then refine the result with follow-up questions or request an interactive dashboard.
OpenAI's examples include diagnosing changes in weekly active users, designing KPI frameworks, and turning monthly metrics into leadership-ready updates. The common thread is that the user starts with the business question rather than with a database query or dashboard-building task.
For companies already invested in Power BI, Tableau, or another supported BI platform, the practical change is the interface between the business question and the finished analysis. OpenAI is trying to make that path conversational while leaving existing data permissions, definitions, and business systems in place.