GuideIntermediate
Building AI Agents Guide
Master the complete lifecycle of building AI agents — from tool calling and function calling patterns to state machines with LangGraph, MCP integration, multi-agent orchestration, and production deployment. Includes testing strategies, evaluation frameworks, and a comparison with Pydantic AI.
- 80
- lessons
- 10
- modules
- English · Spanish
- available in
- Yes
- certificate
- Free
- access
Outcomes
What you'll be able to do
- Understand agent cognitive architecture: perceive-reason-act loop and agent taxonomies
- Build advanced tool use with complex schemas, validation, and error handling
- Master function calling patterns: parallel calls, routing, composition, and structured extraction
- Design agent state machines with LangGraph (StateGraph and Functional API)
- Implement multi-step reasoning: ReAct deep dive, plan-and-execute, reflection and self-correction
- Build memory systems: short-term, long-term, checkpointing, and durable execution
- Integrate external tools with MCP (Model Context Protocol) — the emerging standard backed by Anthropic and Microsoft
- Orchestrate multi-agent systems with supervisor, handoff, subagent, and router patterns
- Test and evaluate agents with trajectory evaluation, golden datasets, and LangSmith
- Deploy agents to production with FastAPI, monitoring, cost control, and error recovery
Before you start
What you need to bring
It's for you if...
- AI Engineers who have completed the LangChain & LangGraph guide and want to build complete agent systems
- Developers who have built basic agents and need production-grade patterns for real-world deployment
- Engineers implementing multi-agent systems at their companies who need orchestration patterns
- Professionals who want to understand MCP and the future of agent-tool integration
- Teams evaluating agent frameworks (LangChain vs Pydantic AI) for production projects
Requirements and materials
- Completed LangChain & LangGraph Guide (#9) or equivalent experience with LangChain v1.2+ and LangGraph v1.0+
- Completed Advanced Prompt Engineering Guide (#10) or equivalent experience
- Advanced Python (async/await, decorators, type hints, Pydantic models)
- Experience building at least one basic agent with LangChain or similar framework
- At least one LLM API key (OpenAI or Anthropic recommended)
- Python 3.11+ installed
Content
The syllabus, module by module
Open any of them to see its lessons.
- 1. Introduction: the world of AI agents
- 2. What an AI agent is (formal definition)
- 3. Cognitive architecture: perceive-reason-act
- 4. A Taxonomy of Agents
- 5. Agents vs Chains vs Workflows
- 6. Framework Landscape 2025-2026
- 7. When to Use (and When NOT to Use) Agents
- 8. Project: A Basic ReAct Agent from Scratch
- 1. Introduction: Agents with memory
- 2. Short-term Memory: Conversation History
- 3. Checkpointing with MemorySaver
- 4. PostgresSaver: Durable Persistence
- 5. Long-term Memory: Cross-Session
- 6. Time-travel Debugging
- 7. Conversation Management and Memory Patterns
- 8. Project: Research Agent v3 — Persistent Memory
- 1. Introduction: Agents Without Tests Are Dangerous Agents
- 2. Unit Testing for Agents
- 3. Integration Testing with a Real LLM
- 4. Trajectory Evaluation
- 5. LangSmith for Agents
- 6. Golden Datasets and Regression Testing
- 7. Benchmarks and Agent Metrics
- 8. Project: Research Agent v6 — Testing and Evaluation Suite
- 1. Introduction: From prototype to production
- 2. Deployment patterns for agents
- 3. Scaling and performance
- 4. Monitoring and Observability in Production
- 5. Error Recovery and Resilience
- 6. Cost Control and Rate Limiting
- 7. Pydantic AI: An Honest Comparison
- 8. Final Project: Research Agent v7 — Production Deployment
Where it fits
This guide is part of something bigger
It's studied inside these programs, with support and dates.
Common questions
What people usually ask
No limit. It's a free guide: come in whenever you like, as often as you like.
No. Modules run from easier to harder, but you can jump to the one you need. Progress is saved per lesson.
Whatever is needed is listed under “What you need to bring”, above. If nothing is listed there, you can start from zero.
In the Club's WhatsApp group, and every two weeks there's a live with an instructor where questions get worked through.
Yes. It's issued automatically once you finish every lesson, with a verifiable code you can share on LinkedIn.
No. This guide is self-paced with no dates. The bootcamp is live, by cohort, with work someone reviews.
Start whenever you like
What students say
These reviews are from enrolled students who completed at least 50% of the course. We moderate reviews only on content grounds (spam, offensive language, personal data), never for being critical or negative.
No approved reviews yet.
Be the first to share your experience!