AI Agents Masterclass for Beginners
From buzzword to builder. AI, LLMs, and Agents — no math, no fluff.
Course content
30 episodesAI in 2026: untangling the buzzwords
Get everyone on the same page. AI demystified without dumbing it down.
- 01lessonsign in
What is AI, really?
Definitions, the Turing test, today's narrow AI vs sci-fi AI, and why "AI" is the most overloaded word in tech.
- 02visualsign in
Narrow, General, Super: the AI spectrum
Where every AI tool you use today actually sits on the capability ladder.
- 03lessonsign in
Machine Learning vs AI vs Deep Learning
Untangling four overlapping terms with a Venn diagram and a simple "when to use which" guide.
- 04quizsign in
Quick check: foundations
Five questions to lock in the basics before we dive deeper.
How LLMs actually work
Demystify the LLM "magic" without math. By the end you can explain why an LLM behaves the way it does.
- 01lessonsign in
Tokens, context windows, and "forgetting"
Why an LLM "forgets" what you said earlier, and how to work with — not against — its memory limits.
- 02lessonsign in
Attention: the secret sauce in plain English
The single mechanism that powers every modern LLM, explained without a single equation.
- 03lessonsign in
Why the same prompt gives different answers
Temperature, top-p, and the dial between deterministic and creative output.
- 04lessonsign in
Prompting 101: zero-shot, few-shot, chain-of-thought
The three patterns that fix 80% of bad LLM outputs.
- 05labsign in
Lab: rewrite three bad prompts
Hands-on: take three real-world bad prompts and turn them into ones that produce great answers.
- 06quizsign in
Quick check: LLM mechanics
Five questions on tokens, attention, sampling, and prompting.
AI Agents: what makes them different
The "aha!" of the course. Why an agent is not just a chatbot.
- 01lessonsign in
Chatbot vs agent: the line in the sand
Why an agent is fundamentally different from a chatbot — even when both are powered by the same LLM.
- 02visualsign in
Anatomy of an agent: brain, eyes, hands
The three parts every agent has: the LLM brain, the tools (hands), and the memory (eyes back into context).
- 03lessonsign in
How agents take action: tool calling explained
Function calling, tool use, structured output — the mechanism that lets an LLM do something instead of just say something.
- 04lessonsign in
Short-term, long-term, and forgetting
Context windows, vector databases, and the trade-offs between remembering everything and remembering well.
- 05lessonsign in
The agent loop: think, act, observe, repeat
The single most important pattern in agentic AI — and the one every modern agent product is a variation of.
- 06flashcardssign in
Lock in the vocabulary
15 essential terms — token, embedding, RAG, tool, MCP, and more — in flashcard form.
- 07quizsign in
Quick check: agent fundamentals
Five questions on what makes an agent different from a chatbot.
Agentic AI patterns
The patterns appearing in every modern agent. ReAct, RAG, planning, MCP — recognise and apply.
- 01lessonsign in
ReAct: Reasoning + Acting
The pattern behind every modern agent — interleaving thought and action until the goal is reached.
- 02lessonsign in
Plan, do, reflect: how agents self-improve
The patterns that take agents from "follows instructions" to "figures it out."
- 03visualsign in
RAG: giving agents your private knowledge
Retrieval-Augmented Generation in one diagram — what it is, when to use it, when not to.
- 04lessonsign in
MCP: the USB-C of AI integrations
The Model Context Protocol — what it is, why it matters, and how it changes the agent ecosystem.
- 05fill insign in
Pattern recognition cloze
Read a description, name the pattern. Five blanks covering ReAct, RAG, planning, tool use, and reflection.
Apply it to your life and work
The section worth taking notes on. Every lesson maps to "5 things to set up tomorrow."
- 01lessonsign in
Personal productivity: 5 agents to set up tomorrow
Email triage, calendar prep, research assistant, meeting notes, learning companion — concrete setups.
- 02lessonsign in
For students and the curious
Studying smarter, research deep-dives, summarising sources, and tutoring loops you can run today.
- 03lessonsign in
For professionals and founders
Meetings, reports, customer support, lead qualification, market research — agent playbooks for work.
- 04lessonsign in
For developers
Coding agents in practice — code review, refactor, test generation, deploy assistance, debugging.
- 05resourcessign in
Tools and resources to start today
A curated list of 15+ tools — free and paid — mapped to skill level and use case.
Hands-on + final assessment
Prove understanding by writing code AND passing a real exam.
- 01codesign in
Build your first agent in Python
~25 lines that define a tool, call an LLM with tool-calling, and run the think-act-observe loop.
- 02lessonsign in
Ethics, pitfalls, and what could go wrong
Hallucinations, bias, prompt injection, privacy. Short, sharp, unmissable for a beginner course.
- 03examsign in
Final exam
15 questions, 70% to pass, 30 minutes. Mix of single and multi-select. On pass: shareable certificate.