There's no shortage of explanations of AI agents online. The harder part is knowing which ones to trust and which are worth your time. This short course does that filtering for you. It pairs the clearest beginner explainers from IBM, a source that knows this material, with a few minutes from me on what each idea means for a team that has to use it at work. It's written for managers and team leaders rather than engineers, and it's free, with nothing gated. Watch it end to end in an hour or so, or take one lesson at a time. Each lesson is one short video and a plain read.
The lessons
Work through them in order, or jump to whichever one you need.
- 01What is an AI agent?What an AI agent actually is, in plain terms: a model put in charge of a goal, working through the steps with tools and memory, not just answering questions.12 min
- 02AI agents vs AI assistantsThe difference between an AI agent, an AI assistant and a chatbot: assistants and chatbots react to you, agents work towards a goal on their own. When to use each.6 min
- 03How an AI agent worksLearn how AI agents actually work inside: the three core parts that turn data into decisions and actions.10 min
- 04How AI agent memory worksHow AI agent memory works: the four types of memory, working, semantic, procedural and episodic, that let an agent learn over time instead of forgetting each session.11 min
- 05How agents use tools (and what MCP is)How agents access the tools and data they need: what Model Context Protocol is, and why it matters when you're building systems that work with real data.4 min
- 06Grounding agents in your own data (RAG)What RAG (retrieval-augmented generation) is, in plain terms: how to ground an AI agent in your own data so it stays accurate and stops making things up.6 min
- 07Multi-agent systems: when agents work as a teamHow multi-agent AI systems work together to solve complex problems. From simple agents to team structures and real-world trade-offs.8 min
- 08Using agents safely: risks and guardrailsThe main risks of AI agents and how to put guardrails on them: shadow AI, data leaks, hallucination laundering, prompt injection, and rogue agents.11 min
- 09The future of AI agents (and what to do now)The future of AI agents, minus the hype: already useful for narrow, structured work, but the messier real-world tasks are years off. How to think about what to try now.13 min
Common questions about this course
Is this AI agents course free?
Yes, completely. Every lesson is free to watch and read, with nothing gated behind a sign-up or a payment at any stage. It pairs nine short IBM videos with a plain-English note from me on what each one means at work, built as a genuinely open resource rather than a lead magnet. So if you want to point your whole ops team at it on a Monday morning, no one hits a paywall halfway through.
Who is this course for, and do I need a technical background?
It is written for managers, team leaders, and anyone at work who is curious about AI but is not an engineer. You do not need a technical background, there is no coding, and there is nothing you need to know in advance. If you run a customer support desk and have never touched a line of code, you can still follow every lesson.
How long does the AI agents course take?
The nine video lessons come to about eighty minutes in total, plus a short read with each one. You can take it two ways:
- End to end: watch the whole thing in an afternoon.
- One at a time: take a single lesson on its own, in order or out of order, so an HR lead could fit one into a lunch break across the week.
What is an AI agent?
An AI agent is a model you put in charge of a goal rather than a single question. A chatbot just answers and stops. An agent plans the steps, draws on the tools and data it needs, and keeps going until the job is done. Think of onboarding a new starter: you hand over the goal, and the agent works through the steps rather than answering one question at a time.
Note: the first lesson explains this in plain terms, and the rest of the course builds from there.