Full course coming soon

Designing Multi-Step AI Workflows for Knowledge Work

7 chapters · 30 lessons

1. 1. Agent Architecture Fundamentals

   4 lessons

   Understand the core components that distinguish agentic systems from single-turn generative AI.

2. 2. Planning and Goal Decomposition

   4 lessons

   Design systems that break down complex goals into executable sub-tasks with clear success criteria.

3. 3. Tool Integration and Execution

   5 lessons

   Connect agents to external systems including APIs, databases, code environments, and business applications.

4. 4. Memory and State Management

   4 lessons

   Implement memory systems that maintain context, track progress, and inform future agent decisions.

5. 5. Human-in-the-Loop Design Patterns

   4 lessons

   Build review checkpoints and approval workflows that balance automation with human oversight.

6. 6. Traceability and Workflow Inspection

   4 lessons

   Make agent decisions auditable through logging, visualization, and structured output formats.

7. 7. Evaluation and Production Deployment

   5 lessons

   Test, monitor, and improve agentic systems in real-world professional environments.