Intermediate

AI Agents

Level 4 moves past single workflows. You design goals, inputs and guardrails, then build a research agent, a sales agent, a support agent and a recruitment agent end to end, before learning how to test them against edge cases and keep them honest in the wild.

6 modules · 8 lessons~80 minutes of reading

Agent loop: think, use a tool, repeat

What separates an agent from a single prompt.

Step 1

Goal

What the user actually wants done

Step 2

Plan

The model picks the next step

Step 3

Tool call

Search, database, calendar, email

Step 4

Finish

Answer or hand off to a human

What you'll cover

Module 1

What Makes Something an Agent

Agent vs automation vs chatbot8 min
Designing goals, inputs and guardrails8 min

Module 2

Research Agent

Step-by-step: build a research agent11 min

Module 3

Sales Agent

Step-by-step: build a sales agent11 min

Module 4

Customer Support Agent

Step-by-step: build a support agent11 min

Module 5

Recruitment Agent

Step-by-step: build a recruitment agent11 min

Module 6

Making Agents Reliable

Testing agents against edge cases9 min
Monitoring and improving agents9 min