Essentials · Episode 2 of 6
Agent and Workflow Design
Most of the quality in an AI workflow lives in how you configure and instruct the agent. The canvas first, then the agent, with two labs to build along with.
Beginner12:11
What You Will Learn
- The canvas: why Start and End are already there, and where the default and fallback models live
- All ten node types in one pass, and why you should start with Agent and add the others only when needed
- Implicit chaining versus explicit references, and why a vague input schema is the most expensive day-one mistake
- Everything you set on an agent, and why its type is derived from what you attach rather than picked
- Four rules for writing instructions, and how the model for a step is chosen
- Lab 1 and Lab 2: build a first workflow, then add structured output and test the unhappy paths