01 December 2026 14:00 - 14:30
From simple loops to long-running agents
When is a simple agent loop enough, and when should you give an agent more time or a teammate?
Drawing on examples from OpenAI, weāll explore how the design changes with the job, from diagnosing a production slowdown to coordinating a longer research project.
Weāll look at what engineers need to build so agents can test their ideas, pick up where they left off, and check each otherās work.