Build Hour: Workspace agents in ChatGPT

By OpenAI

Categories: AI, Product

Summary

OpenAI's workspace agents enable teams to automate complex multi-system workflows like meeting prep and vendor review directly in ChatGPT and Slack. Built on Claude with memory and persistent background execution, they're positioned between personal Codex agents and custom SDK implementations for enterprise teams.

Key Takeaways

  1. Workspace agents handle long-running tasks spanning multiple systems with file, code, and tool access—OpenAI's marketing team uses them to convert product briefs into websites pulling from Google Docs and code simultaneously.
  2. Three-tier agentic product strategy: workspace agents for team/shared work, Codex for individual personal agents, Agents SDK for custom product integrations—each solves different deployment and scale needs.
  3. Meeting prep agent reduces manual evening work by automating calendar checks, web/Drive research, and brief creation daily—real enterprise use case showing ROI on agent development time.
  4. Agents have persistent memory and can be guided through conversation to improve over time—enabling iterative refinement without rebuilding, similar to fine-tuning workflows.
  5. Available in research preview across ChatGPT Business, Enterprise, Edu, and Teachers plans—indicates enterprise-first go-to-market with compliance/admin controls built in.

Topics

Transcript Excerpt

Hello and welcome back to build hours. I'm Victoria from product marketing and I'm here with Christina and Hojun. >> Hi, I'm Christina and I work on the engineering team for workspace agents. >> Hi, I'm Hojun. I'm on the solutions engineering team. >> Awesome. And today we're going to show you workspace agents in chatt. So if this is your first build hour, um build hours is all about helping teams get more practical value from open a products with real examples and tips you can use after the ses...