Course · 9 chapters
Agent framework
Build production multi-agent systems with the OpenAI Agents SDK and Claude Agent SDK, the engine behind Claude Code. 9 chapters for working engineers.
What you'll be able to do
- Master the OpenAI and Anthropic APIs
- Build multi-agent systems with the OpenAI SDK
- Ship autonomous agents with the Claude Agent SDK
- Add guardrails, handoffs, and tracing
- Test non-deterministic agents in CI
- Compare LangGraph, CrewAI, and AutoGen
What's inside
- 1Agent Frameworks: parti da qui
Un orientamento di 12 minuti sullo skill path Agent Frameworks — perché esiste, cosa imparerai a fare, come si collegano i capitoli e da dove iniziare.
- 2Agent framework: il panorama
Naviga lo stack che va dall'API all'agent, lato OpenAI e Anthropic — capisci i livelli prima di costruire.
- 3OpenAI API: mastery
Padroneggia sia la Responses API sia Chat Completions — capisci quando usare quale e come migrare.
- 4Anthropic API & SDK
Padroneggia la Messages API, il tool use, l'extended thinking e il prompt caching — costruisci con Claude partendo dai principi fondamentali.
- 5OpenAI Agents SDK
Costruisci sistemi multi-agente con guardrail, handoff e tracing — il framework di OpenAI per l'AI autonoma.
- 6Claude Agent SDK
Costruisci agenti autonomi con il framework di Anthropic — lo stesso motore che alimenta Claude Code, ma come libreria che controlli tu.
- 7Workflow Agent in Produzione
Porta in produzione sistemi multi-agent che si integrano con strumenti reali, scalano nei container e verificano il proprio lavoro.
- 8Test e valutazione degli agent
Rilascia agent verificabili — mock dei tool, traccia delle decisioni, golden dataset e gate sui deploy in base agli eval.
- 9LangGraph e agenti multi-framework
Leggi, esegui il debug e porta in produzione su LangGraph, CrewAI e AutoGen — i tre framework che incontrerai nel codice dei clienti.
Frequently asked questions
- What will I be able to build after this Agent Frameworks course?
- You go from raw API calls to production multi-agent systems. You build directly with the OpenAI Agents SDK and the Claude Agent SDK (the engine behind Claude Code), wire in tool use, multi-agent handoffs and guardrails, test agents in CI/CD, deploy production workflows, and learn to read and debug the LangGraph, CrewAI, and AutoGen frameworks you encounter in real customer code.
- Who is this course for?
- It is built for software engineers at a practitioner level who want to build, test, and ship AI agents. It suits backend, full-stack, and AI engineers moving from raw API calls to production agent systems.
- Do I need coding experience to take this course?
- Yes. This is a practitioner-level engineering path that has you working directly with the OpenAI and Anthropic SDKs, so you should be comfortable reading and writing code before starting.
- How long is the course and is there a certificate?
- The path has 9 chapters, including a Start Here orientation, the landscape of agent frameworks, the OpenAI and Anthropic APIs and SDKs, production workflows, testing and evaluation, and multi-framework agents. You earn a certificate of completion when you finish the path.
- Which agent frameworks does the course cover?
- It covers the OpenAI Agents SDK and Anthropic's Claude Agent SDK in depth, plus the OpenAI Responses API, the Anthropic Messages API, and reading and debugging agents built with LangGraph, CrewAI, and AutoGen.
Earn a certificate
Complete all chapters to receive your certificate of completion.