Students will:
- Understand agents and identify appropriate use cases.
- Translate work into agent workflows.
- Control agent behavior and actions.
- Manage agent memory and knowledge.
- Design safe, secure agents with human oversight.
- Test, evaluate, and monitor agent performance.
Course Outline:
Lesson 1: From Manual Work to Agentic Systems
Agentic AI is reshaping modern work, but its power depends on how well you understand orchestration and responsible oversight. In this first module, you will take a short pre-assessment to gauge your starting point and learn how the course is organized. You will study the foundations of what defines an agent, determine when an agentic approach is appropriate for a task, and learn how to make work legible to an AI system through structured workflows.
Lesson 2: Controlling Agent Behavior, Actions, and Memory
This module focuses on the high-level levers that define the power and boundaries of an agent. Through simulations, you will learn to oversee how agents are guided by specific instructions, how they connect to the real world through tools, and how they manage information over time to ensure privacy and consistency.
Lesson 3: Safety, Oversight, and Agent Performance
This module focuses on the principles of responsible oversight and performance monitoring. You will learn to apply least-privilege principles to agent permissions, identify high-risk points that require a human-in-the-loop, and use monitoring signals to detect when an agent’s accuracy or behavior changes over time.
