The skills the work actually rewards
These are the core skills an AI evaluator uses every day, taught from the ground up. No coding, no AI background, and no jargon to get started.
Six skills you build first
Pulled straight from the curriculum, focused on what the work actually asks of you. No prerequisites to start.
Designing evaluation rubrics
Write atomic, objective scoring criteria from scratch, the way the platforms expect. The skill behind the most demanding evaluation work.
Defining the ideal response
Describe exactly what a perfect answer looks like for a prompt, the starting point for rating and ranking anything an AI produces.
Comparing and ranking responses
Apply a clear method to decide which of two answers is better, and defend the call. The core of how models learn from human feedback.
Spotting errors and hallucinations
Judge accuracy and instruction-following, and catch the confident mistakes models make, across text, code, and images.
Writing defensible justifications
Explain every rating with evidence so a reviewer can follow exactly why you scored what you did. Expected on every platform.
Fact-checking and citations
Verify claims quickly and tell strong sources from weak ones, the difference between a passing and a failing evaluation.
Want the full picture?
These six are the foundation. The full certification runs to 24 modules across rubric design, response evaluation, safety, citations, and platform readiness.