WBF Academy
AI Agents Mastery

AI Agents Mastery

Intermediate

Build production-grade AI agents from first principles. Master the ReAct loop, tool use, memory, planning, RAG, and multi-agent orchestration. Code examples use Anthropic's Claude SDK; patterns port to any LLM provider. Frameworks covered: LangGraph, CrewAI, AutoGen.

📋 5 tracks ❓ 350 questions 💡 35 tips 🎬 15 videos ⏱ ~12h
▶ Start watching

Course content

What is an AI Agent 6:16

What is an AI Agent

⏱ 6:16

▶
The ReAct Loop 6:26

The ReAct Loop

⏱ 6:26

▶
Building Your First Agent with Anthropic SDK 6:59

Building Your First Agent with Anthropic SDK

⏱ 6:59

▶
🧠

Practice questions

0/70 answered

Start
Tool Calling Deep Dive 6:26

Tool Calling Deep Dive

⏱ 6:26

▶
Short-term Memory 6:28

Short-term Memory

⏱ 6:28

▶
Long-term Memory with Vector Stores 6:47

Long-term Memory with Vector Stores

⏱ 6:47

▶
🧠

Practice questions

0/70 answered

Start
Task Decomposition with Planner-Executor 6:07

Task Decomposition with Planner-Executor

⏱ 6:07

▶
RAG for Agents 6:19

RAG for Agents

⏱ 6:19

▶
Self-Correction with Reflection 6:15

Self-Correction with Reflection

⏱ 6:15

▶
🧠

Practice questions

0/70 answered

Start
Multi-Agent Patterns 6:19

Multi-Agent Patterns

⏱ 6:19

▶
LangGraph Tour 6:33

LangGraph Tour

⏱ 6:33

▶
CrewAI and AutoGen Compared 6:59

CrewAI and AutoGen Compared

⏱ 6:59

▶
🧠

Practice questions

0/70 answered

Start
Evals and Observability 6:22

Evals and Observability

⏱ 6:22

▶
Guardrails and Safety 7:03

Guardrails and Safety

⏱ 7:03

▶
Deploy and Cost Control 7:06

Deploy and Cost Control

⏱ 7:06

▶
🧠

Practice questions

0/70 answered

Start

Certification Exam

✦
🏆

Certification Exam

AI Agents Mastery

❓ 40
Questions
⏱ 120m
Time Limit
% 75%
To Pass

🔒 Unlock the exam

🎬 Watch 80% of the videos 0% / 80%
❓ Answer 80% of the practice questions 0% / 80%

Complete the requirements above to unlock the exam

🏆

Certification Exam

AI Agents Mastery

#

40 Questions

All difficulty levels

⏱

120 Minutes

Auto-submits when time expires

%

75% to Pass

Earn your certification badge

→

No Going Back

Once you answer, you move forward

Begin Exam

Tips

See all
MultiAgent Intermediate

Agents talk through state, not chat

A shared object beats free-form messages
Performance Intermediate

Cache the boilerplate, pay for the diff

Prompt caching cuts cost up to 90%
Production Intermediate

Route by request, not by hope

Cheap model for easy, big model for hard
Evals Beginner

Build the eval set before the agent

No dataset = no progress, only opinions