How data science teams use ChatGPT Work

ChatGPT Work allows data science teams to accelerate the creation of analysis assets by synthesizing scattered inputs into structured first drafts. This capability enables teams to move from raw data and business questions to validated deliverables more efficiently.

Accelerating Analysis Asset Creation

ChatGPT Work streamlines the process of turning disparate data sources into review-ready analysis. By integrating various inputs, the tool helps data scientists assemble a first draft of their deliverables, which typically includes:

• Visualizations: The generation of charts to represent data findings.
• Contextual Documentation: The inclusion of caveats and source links to ensure transparency.
• Review Frameworks: The creation of review questions to facilitate validation by the team.

Supported Data Inputs for Data Science

To generate these analysis assets, ChatGPT Work can process a variety of scattered inputs, including:

• Dashboards: Existing data visualizations and summaries.
• Metric Definitions: The specific parameters and logic used to define key performance indicators.
• Metric Exports: Raw data exports from internal systems.
• Experiment Notes: Documentation regarding the testing and hypotheses of data experiments.
• Business Context: General organizational goals and background information necessary to interpret the data.

Integration and Availability

These workflows were previously available in the Codex app and have now been integrated into ChatGPT Work. Users can access these capabilities via chatgpt.com or the ChatGPT desktop application.