VIDEO COURSE · intermediate
Build Useful AI Agents & Automations
Design a focused agent workflow with clear tool boundaries, approvals and recovery paths.
This course is in development. The curriculum is a planned outline; recorded lessons and PDF slides are not yet available.
What you’ll learn
- Choose an appropriate automation task
- Define agent permissions and tool contracts
- Evaluate failure and recovery
Before you start
- Comfort with basic UX methods
- A practical project to apply the lessons
Planned curriculum
Choose a useful task
- Separate automation from product value
Practice: separate automation from product value. Capture your decisions and evidence in a project worksheet.
Related prompts → - Map inputs, outputs and human decisions
Practice: map inputs, outputs and human decisions. Capture your decisions and evidence in a project worksheet.
Related prompts →
Build the workflow
- Define tools and permission boundaries
Practice: define tools and permission boundaries. Capture your decisions and evidence in a project worksheet.
Related prompts → - Add checkpoints and recovery
Practice: add checkpoints and recovery. Capture your decisions and evidence in a project worksheet.
Related prompts →
Evaluate reliability
- Test representative and adversarial cases
Practice: test representative and adversarial cases. Capture your decisions and evidence in a project worksheet.
Related prompts → - Measure quality, cost and completion
Practice: measure quality, cost and completion. Capture your decisions and evidence in a project worksheet.
Related prompts →