Builder Bootcamp: Agents

In this session, you’ll learn how production-grade agents are structured by walking through a customer support agent built with the OpenAI Agents SDK. We’ll cover how agents use tools to look up orders, answer FAQs, cancel orders, start complaints, route requests, apply guardrails, and generate outputs that can be evaluated.

We’ll connect the core concepts from the session, including tools, instructions, handoffs, routing, guardrails, tracing, and evals, to a concrete Python implementation. By the end, you’ll understand how to move from a single-agent workflow to more advanced patterns like agent-as-tool delegation, triage routing, guardrail tripwires, and model-graded evaluation.