Evaluating AI Output
Theme: Using AI toolsLevel: Intermediate
How to tell whether a prompt, a model or an AI tool is really good — define what "good" means, understand what makes a solid test set, score answers with checks, people and AI judges, and catch problems when something changes. No code needed.
Before this course
This course builds on:
Lessons
Why "it looked fine" is not enough
Defining what good means
What makes a good test set
Scoring: checks, people and AI judges
Comparing versions and catching regressions
Evaluating in real use