Learn AI · Guided path 4 of 6

What makes it an agent

When you finish: You can tell a model, a chatbot, a workflow and an agent apart, and describe the loop an agent runs.

About 25 minutes · 9 stops, 6 of them ideas from the map · progress is kept on this device ·

A requester brings a damaged lamp to a workshop. A software operator works at a cobalt context tray while an owner and a reviewer oversee the boundary.

Outline operator: software metaphor · ⏸ Review boundary

Read the concept diagram

Start at stop 1 All paths

The stops, in order

Each stop is one page. Read it, picture it, mark it, move on.

Ideas open on their own page with three depths. Atlas stops open the interactive lesson at the right step and count as done when you follow the run to its end. Sections count when you have read them.

1Model, chatbot, workflow, agentFour kinds that get called the same thing.section
2AI AgentsAI that doesn't just answer but takes steps: searching, using tools, and checking its own work toward a goal.idea
3Tools and Function CallingLetting a model ask your software to do something: search, read a file, create a ticket, run a query.idea
4MCPModel Context Protocol: a standard way to plug tools and data sources into any AI model, so each integration is built once.idea
5MemoryHow an AI keeps track of things beyond the current conversation.idea
6Retrieval (RAG)Letting an AI look things up in trusted documents before it answers.idea
7The agentic loopFour stations, one exit.section
8Follow one request through a governed systemThe flagship lesson, start to verified completion.atlas lesson
9Multi-agent SystemsSeveral agents with different jobs working together, usually with one coordinating the others.idea

← Path 3 Path 5: Running it safely →