This intermediate path teaches you how to design and implement GPT-5-powered agents from scratch in Python. You will work with core agentic workflows, including task decomposition, routing, iterative tool use, and response orchestration. You will learn how to extend agents with custom tools and functions so they can interact with external systems and complete multi-step tasks. The path also covers collaboration patterns for specialized agents, including delegation, coordination, and orchestration across multiple roles. The later courses focus on scaling agentic systems with asynchronous programming. You will practice running concurrent API calls, parallel tool operations, and multiple agents at the same time to support higher-throughput workloads.
