Home›Agentic AI Tutorial
Free course
Agentic AI Tutorial
Learn to build AI agents, from one simple loop to a production system, step by step, in simple words. Every idea is shown with animations, real code and real model runs.
What you will learn
- The agent loop, and when not to use an agent
- Tool calling, structured output, memory and agentic RAG
- Workflows: chains, routers, orchestrators and evaluators
- LangGraph: state, checkpoints, human approval and time travel
- Multi-agent teams, MCP servers and A2A
- Security: prompt injection, the lethal trifecta and sandboxes
- Coding agents, evals, tracing and production rollout
- A full capstone agent, quizzes and interview preparation
Course contents
Every lesson has its own page with the explanation, code you can copy, the real output, screenshots from the video, and a link that jumps straight to that moment in the video.
1. What Agents Are
2. Agents and Models
3. Talking to Code
- Structured Output12:34
- Tokens, Context, and Cost14:30
- Choosing a Model16:12
4. Tools
5. Real Tool Use
- Real Function Calling22:21
- Designing Great Tools24:27
- When Tools Fail26:54
6. Parallel Tools and Limits
7. Workflows
- Prompt Chaining34:56
- Routing36:49
- Parallelism38:33
8. Workflows Two
9. Memory
- Conversation Memory44:53
- Summarising the Past46:20
- Long-Term Memory48:02
10. Knowledge
- Searching by Meaning49:56
- Agentic RAG51:41
- When Memory Goes Wrong53:21
11. Planning and Reasoning
- Thinking Step by Step55:16
- Plan and Execute57:01
- Reflection58:47
12. Finishing Well
- Knowing When to Stop1:00:34
- Mini Project: A Research Agent1:01:59
13. LangGraph Basics
- Why Graphs?1:03:50
- Your First Graph1:05:17
- Conditional Edges1:06:32
14. LangGraph Agents
- An Agent in LangGraph1:08:08
- State and Reducers1:09:57
- Checkpoints1:11:37
15. LangGraph Power
- Human in the Loop1:13:37
- Edit and Approve1:15:32
- Streaming1:17:19
- Time Travel1:18:34
16. Multi-Agent Systems
- Why Many Agents?1:20:07
- The Supervisor Pattern1:21:47
- Handoffs1:23:30
17. Team Design
- Sharing Context1:25:07
- Critics and Debate1:26:50
- How Agent Teams Fail1:28:28
18. Model Context Protocol
- The Problem MCP Solves1:30:17
- Building an MCP Server1:32:02
- MCP in the Real World1:34:00
19. Standards and Frameworks
- Agents Talking to Agents1:35:44
- A Tour of Agent Frameworks1:37:46
- Instructions as Files1:39:43
20. Security One
- Prompt Injection1:41:29
- The Lethal Trifecta1:44:03
- Least Agency1:46:00
21. Security Two
- Guardrails in Layers1:47:50
- Sandboxing Code1:50:04
- The OWASP Top Ten for Agents1:52:08
22. Agents That Build
- A Coding Agent1:53:59
- Working with Coding Agents1:55:59
- Code as a Tool1:57:27
23. Evals
- Why Evals Matter1:59:20
- A Golden Test Set2:00:56
- Testing the Path2:02:33
- Models as Judges2:03:53
24. Evals and Tracing
- Evals in Your Workflow2:05:40
- Tracing2:07:25
- Agent Metrics2:08:59
25. Production
- Faster and Cheaper2:10:43
- Serving an Agent2:12:25
- Rolling Out Safely2:14:38
26. Real-World Agent Design
- Designing for People2:16:28
- Voice Agents2:18:09
- Background Agents2:19:41
27. Agents at Work
- Agents and Databases2:22:11
- Long-Running Tasks2:23:47
- Automatic Compaction2:25:16
- Many Customers, One Agent2:26:12
28. Lessons from the Field
- When Agents Went Wrong2:27:45
- Build or Buy?2:29:59
- Where Agents Are Heading2:31:30
29. Review: Real-World Design
30. Myths
- Agent Myths2:39:39
31. Capstone
- Capstone: Designing Juno2:43:09
- Capstone: Juno's Tools2:44:50
- Capstone: Juno's Graph2:46:03
- Capstone: Attacking Juno2:47:36
- Capstone: Proving It Works2:49:07
32. Quizzes
- Quiz: Foundations2:50:43
- Quiz: Graphs and Teams2:55:54
- Quiz: Safety and Evals3:01:02
33. Mastery
- Interview Questions3:06:13
- Your Cheat Sheet3:08:03
- Your Roadmap3:08:53