Meet the Data Agent in ChatGPT Work

**TL;DR:** ChatGPT's Data Agent integrates trusted data and team expertise into daily workflows, transforming business questions into actionable plans with shared dashboards, alerts, and consistent metrics. ## From Business Question to Actionable Plan A team member receives a management query about a new product launch via mobile during their commute. Using ChatGPT Work, the Data Agent accesses trusted company data and existing site context to begin analysis. By the time the user reaches their desk, the analysis is ready. It identifies a key opportunity: returning creators are driving growth, but the conversion funnel shows only about 16% of users who gain access actually start creating content. The insight is clear: more customers need help taking their first step. ## Deep Dive with Data Plugins The user employs data plugins to examine the evidence behind the analysis. This includes reviewing metric definitions, underlying data, and the SQL queries that generated the report. The data team can audit the answer generation process to ensure it addresses the correct business question. The user then checks existing reports in Slack conversations and Google Drive for clues about why customers aren't creating sites. ## Synthesizing Insights and Planning Next Steps Feedback indicates customer uncertainty about how to start. This leads to a concrete recommendation: provide in-app guidance for users on how to publish a site. The user then asks ChatGPT to help assign owners and define follow-up steps. After reviewing the plan, it is shared in Slack so the right people can take action with clear responsibilities and shared progress metrics. ## Creating a Shared Team Dashboard The data plugin transforms this analysis into a dashboard. This provides product owners, customer success teams, and data teams with a shared view based on the same trusted source and definitions. The dashboard answers recurring weekly questions: Are more customers adopting the product? Are creators returning? Which customer segments need more help? Teams can self-serve these questions, freeing the data team for deeper investigations. ## Setting Up Automated Alerts To maintain visibility, the user sets up an alert. They use ChatGPT to define the rules: what constitutes a significant drop in weekly creator adoption, when to check, and who should be notified. These details are reviewed before activation. With a refresh option, the shared dashboard stays updated throughout the launch period. This provides a continuous way to spot changes, align on priorities, and decide when to adjust plans. ## Building a Consistent Data Context Layer To ensure consistent data across the organization, the user leverages the data context layer. The data team built this foundation using the Data Agent. It helps synthesize knowledge from data warehouses, analytical conversations, and documentation into a single source of truth. This context defines business terms (e.g., "visiting a site" vs. "creating a site"), specifies which sources and activities are counted, sets reporting periods, and standardizes comparisons. It ensures unified metric definitions organization-wide, so any data question is answered with consistent, trusted results based on the same business context. ## Putting Trusted Data into Practice Data plugins bring trusted analytics, shared dashboards, and consistent business context into daily work. This makes the data team's expertise accessible to everyone who needs to make decisions, embedding trusted data directly into the workflow. Source: [Meet the Data Agent in ChatGPT Work](https://www.youtube.com/watch?v=VMgEQ7ym9WU)