OpenAI has launched a Data agent within ChatGPT Work designed to allow business users to query data and generate interactive dashboards using natural language. The agent integrates directly with popular data platforms including Amazon Redshift, Google BigQuery, Databricks, Snowflake, and MongoDB, alongside documents from Google Drive and SharePoint. By leveraging semantic layers such as dbt and Databricks Genie Ontology, the tool interprets organizational terminology, custom calculations, and security permissions.
According to OpenAI, enterprise administrators maintain control over available data sources and user permissions down to table, row, and column restrictions. Beyond generating visualizations, the agent can recommend next operational steps, send updates through Slack or email, and execute approved workflows in external analytics tools such as Tableau, Power BI, and ThoughtSpot.
OpenAI reports that the tool is already widely used internally across its GTM and product teams, while organizations like NTT Data and Thermo Fisher have adopted it through an early Alpha program. Administrators can deploy the agent through the ChatGPT Work Plugins directory.
Why it matters
Enables non-technical enterprise employees to perform complex data analysis without writing SQL queries or relying on specialized BI teams.
Integrates directly into existing data stacks and semantic layers, maintaining strict enterprise permission controls and governance.
Expands LLM utility from passive text generation to active workflow execution across popular enterprise productivity platforms.
Source: openai.com
