OpenAI SDK实战：从零构建智能AI代理系统

本课程教你用Python和OpenAI Agents SDK构建AI智能体，涵盖工具调用、RAG、MCP及多智能体协作，助你掌握从调试到部署的全栈开发技能。

原始标题：Build AI Agents with OpenAI SDK: Build Real Agentic AI

该课程是一门面向开发者的 AI 智能体（AI Agents）实战进阶指南，旨在指导学员使用 Python 和最新的 OpenAI Agents SDK，将大语言模型从被动的对话框升级为能自主思考、调用工具并解决复杂任务的主动执行者。课程从零基础构建智能体开始，核心涵盖函数调用（Tool Calling）、利用 OpenAI Tracing 进行工作流链路调试，以及融合检索增强生成（RAG）来扩展智能体的知识边界。

同时，课程聚焦当前最前沿的分布式 AI 架构，重点传授 MCP（模型上下文协议） 的应用以及多智能体协同（Multi-Agent Workflow） 的设计。学员将学习如何让多个具备不同专业分工的 Agent 相互沟通、组队协作，从而攻克企业级复杂业务场景，最终掌握从逻辑设计、生产调试到多 Agent 调度部署的全栈智能体开发能力。

Published 8/2026
MP4 | Video: h264, 1920×1080 | Audio: AAC, 44.1 KHz, 2 Ch
Language: English | Duration: 2h 7m | Size: 965.68 MB

A Hands-On Guide to Building, Debugging, and Deploying Intelligent AI Agents with Tools, RAG, MCP & Multi-Agent Workflow

What you’ll learn
Build AI agents from scratch using the OpenAI Agents SDK and Python.
Create real-world agent capabilities using function tools and tool calling.
Use OpenAI Tracing and observability to debug agent workflows, LLM calls, and tool calls.
Build advanced agentic AI applications using RAG, MCP, and multi-agent architectures.

Requirements
Basic Python programming knowledge is recommended.
Basic understanding of APIs and JSON will be helpful, but is not required.
An OpenAI API key to run the hands-on examples. API usage may incur a small cost.
No prior experience with AI agents, Agentic AI, RAG, MCP, or multi-agent systems is required.

Description
Want to move beyond basic ChatGPT prompts and learn how to buildreal AI agents?

Welcome toMaster AI Agents with OpenAI SDK: Build Real Agentic AI— a practical, hands-on course where you’ll learn how to build intelligent agentic applications using theOpenAI Agents SDK.

We’ll start from the fundamentals and gradually build more capable AI agents. You’ll learn how to create agents with the OpenAI SDK, provide instructions, execute agents using Runner, and give them real capabilities throughtool calling.

We’ll exploreOpenAI Tracing and observability so you can see exactly what happens inside your agent — including LLM calls, tool calls, and execution flows.

As we progress, we’ll expand our agents withRAG, MCP, and multi-agent architectures, building toward real-world agentic AI applications.

This course is designed to be practical and beginner-friendly, with working examples, hands-on coding, quizzes, and clear explanations throughout.

This course is designed to be practical and beginner-friendly, with working examples, hands-on coding, quizzes, and clear explanations throughout. You’ll learn by building rather than simply watching theory, allowing you to understand not onlyhow agents work, but alsowhy different architectural decisions matter.

By the end, you won’t just understand what agentic AI is —you’ll know how to build it.

Who this course is for
Python developers who want to learn how to build AI agents and Agentic AI applications.
Software developers and engineers looking to integrate LLMs into real-world applications.
AI/ML engineers who want hands-on experience with the OpenAI Agents SDK.
Beginners in Agentic AI who want a practical, step-by-step introduction to agents, tools, RAG, MCP, and multi-agent systems.