AI Agents (Agentic AI)
Learn agentic AI the way teams ship it: planners, tool routers, memory, human approvals, and observability. Practice with popular frameworks and patterns, then deliver a mentor-reviewed capstone agent that completes a real multi-step workflow—not a single prompt that hallucinates a plan.
Model agent loops: plan → act → observe → reflect
Connect tools (search, APIs, code) with permission boundaries
Add memory, checkpoints, and human-in-the-loop approvals
Ship a capstone agent with traces and failure handling
Overview
What you'll learn
- Model agent loops: plan → act → observe → reflect
- Connect tools (search, APIs, code) with permission boundaries
- Add memory, checkpoints, and human-in-the-loop approvals
- Ship a capstone agent with traces and failure handling
Curriculum outline
Weekly milestones
Week 1
View topicsAgent vs chatbotReAct patternTool contracts
Week 1
View topicsWeek 2
View topicsPlanner promptsSub-tasksState machines
Week 2
View topicsWeek 3
View topicsTool routingAPI toolsError recovery
Week 3
View topicsWeek 4
View topicsMemory layersSession vs long-termSummarization
Week 4
View topicsWeek 5
View topicsMulti-agent patternsHandoffsSupervisor agents
Week 5
View topicsWeek 6
View topicsObservabilityTracesEvaluating agents
Week 6
View topicsWeek 7
View topicsCapstone scopingBuild sprintMentor review
Week 7
View topicsWeek 8
View topicsSafety reviewDemoDeployment notes
Week 8
View topicsWho this is for
A best-fit for Intermediate
This track is built for learners who want design multi-step agents with tools, memory, and orchestration—ship an agent capstone.. You'll learn with structured milestones and mentor feedback so you always know how to progress.
Mentor support
You'll receive regular touchpoints with mentors for review, guidance, and decision-making coaching. The goal is not just completion—it's building proof.