anonymous · 14:35 4/6/26 · ORA·techanonymous · 14:35 4/6/26 · ORA·techanonymous · 14:35 4/6/26 · ORA·techanonymous · 14:35 4/6/26 · ORA·techanonymous · 14:35 4/6/26 · ORA·techanonymous · 14:35 4/6/26 · ORA·tech

AI AGENTS COURSE

Learn AI Agents — A Structured Course

From “what is an agent” to the core design patterns (Tool Use/ReAct, Planning, Metacognition) — with interactive simulations. Adapted from Microsoft’s AI Agents for Beginners curriculum.

17 lessons · 5 stages 3 interactive simulators Middle plan and up

Prerequisites:you should know the basics of LLMs & prompting first. If not, take the AI Fundamentals course then come back — it’ll be much easier.

Key terms to remember

Agent — A system that uses an LLM to decide its own steps to reach a goal (not just to answer).

Tool / Function calling — The mechanism by which an agent calls functions/APIs to act and fetch real data.

ReAct — A pattern that alternates Reasoning and Acting in a loop.

RAG — Retrieval-Augmented Generation — retrieve relevant documents before generating an answer.

Orchestration — Coordinating multiple agents/steps to work together.

MCP / A2A — Standard protocols for agents to connect to tools (MCP) and talk to each other (A2A).

Note: we keep the English terms (agent, tool, ReAct…) — translating them often makes things harder to follow.

Introduction to AI Agents

Understand what an agent is, when to use one, and what to build it with.

Lesson 1.1What is an AI Agent?

After this lesson you will

Tell an “AI agent” apart from an ordinary chatbot/LLM.
Name the 4 core components of an agent: goal, brain (LLM), tools, and loop.

An AI agent is a system that uses a language model (LLM) as its “brain” to decide on its OWN which steps to take to reach a goal — rather than just answering a question. The key difference from a chatbot: an agent can act (call tools, read data, iterate) and adjust itself based on the results.

The four components of an agent

A minimal agent has four pieces that work together inside a loop.

Goal: what the user wants to achieve, not a fixed command.
Brain (LLM): reasons, plans, and picks the next action.
Tools: functions/APIs the agent calls to “touch” the real world (web search, read a DB, send mail).
Loop: think → act → observe the result → repeat until done.

How an agent differs from a chatbot

A chatbot takes a question and returns an answer. An agent takes a goal, breaks it into multiple steps on its own, calls tools, checks the results, and only then concludes.

Chatbot: 1 input → 1 output. Agent: 1 goal → many self-directed steps.
An agent has a “working memory” and can change course when it hits unexpected results.
Autonomy is a spectrum: from “suggest for a human to approve” to “fully self-running”.

Common pitfalls

Assuming an LLM plus a few APIs makes an agent — without the “observe & correct” loop it is just a script.
Granting too much autonomy from the start; begin at the “propose, human approves” level.

Tip: When starting out, ask “does this task need multiple steps + tool calls + self-correction?”. If not, a good LLM prompt is usually cheap and enough — don’t build an agent for a one-step job.

Unlock the entire AI Agents course

Lesson 1.1 is a preview. Unlock the remaining lessons + 3 interactive simulators.

Requires the Middle plan or higher

Already have an eligible plan? Sign in to unlock right away.

Core Design Patterns

The durable patterns that every framework revolves around.

Knowledge & Memory

Give the agent the right data at the right time — and remember what matters.

Multi-Agent & Protocols

Multiple agents collaborating, connecting tools, and acting in the real world.

Trust · Security · Operations

Take agents to production safely and reliably.

Coming soon

Deploying agents at scale (Scalable) Coming soon

Building local-running AI agents Coming soon