Full course coming soon

Designing Agentic AI Workflows: Planning, Tools, and Governance

7 chapters · 32 lessons

1. 1. From Single Response to Multi-Step Action

   4 lessons

   Understand how agentic systems differ from chatbots by planning tasks, calling tools, and revising based on results.

2. 2. Tool Calling and External Context Retrieval

   5 lessons

   Connect agents to APIs, databases, and knowledge sources so they can act on real data and systems.

3. 3. Designing Constraints and Permissions

   4 lessons

   Build safety into agents through access controls, guardrails, and explicit boundaries on what actions are allowed.

4. 4. Human-in-the-Loop and Approval Workflows

   5 lessons

   Decide where humans stay in control and how to design supervised agents for high-stakes environments.

5. 5. Evaluating Agent Performance and Failure Modes

   5 lessons

   Test agents for accuracy, safety, and reliability using real scenarios and edge cases.

6. 6. Real-World Agent Use Cases and Process Integration

   5 lessons

   Apply agent design to customer support, operations, and software workflows with domain-specific considerations.

7. 7. Building a Portfolio-Ready Agent Project

   4 lessons

   Create a small, documented agent that automates a real task and demonstrates design decisions.