Skills you can demonstrate, not videos you watched.
Every capability below is backed by a lesson, a game, or a Practice Bay rubric — and a passing score is what unlocks the certificate. Filter by 4D dimension and level.
Dimension
Level
Discover
Foundational
Frame an AI opportunity for HR
Translate a vague exec ask into a scoped, defensible HR use case.
Proof · 1-page use-case brief, graded against scoping rubric.
2 lessons
Discover
Intermediate
Build the business case for AI in HR
Quantify cost, time, and risk reduction in a way a CFO will sign off on.
Proof · ROI memo with assumptions, sensitivity, and risk register.
1 lesson
Develop
Foundational
Prompt without leaking PII
Get useful output from an LLM while keeping POPIA, GDPR, and IP safe.
Proof · Before/after prompt pair, rubric-scored on safety and clarity.
1 lesson
Develop
Intermediate
Apply 5 reusable HR prompt patterns
Move beyond one-shot prompting — chain, critique, and constrain.
Proof · Five patterned prompts, each graded against its dimension's rubric.
1 lesson
Develop
Advanced
Design an evaluation rubric with AI
Co-author a defensible rubric you can stand behind in a CCMA hearing.
Proof · Original rubric with anchored examples at levels 1, 3, and 5.
1 lesson
Deliver
Foundational
Write a bias-free Job Description
Skills-first, inclusive, specific. Pass the JD Rewriter rubric.
Proof · JD scored ≥ 4/5 by the JD Rewriter game.
1 lesson1 game
Deliver
Intermediate
Build a structured interview kit
Competency-mapped questions, scoring guide, and panellist brief in 20 min.
Proof · Interview kit graded against the structured-interview rubric.
1 lesson
Deliver
Foundational
Rewrite vague feedback as SMART
Turn 'be more proactive' into Specific, Measurable, Achievable, Relevant, Timely.
Proof · Feedback rewrite scored ≥ 4/5 by the SMART Feedback game.
1 lesson1 game
Defend
Intermediate
Stand up AI governance you'll actually use
Lightweight policy, register, and review cadence — not a 40-page binder.
Proof · One-page AI policy + 1-page use-case register template.
1 lesson
Defend
Intermediate
Apply POPIA to an AI workflow
Lawful basis, minimisation, transfer, retention — for every HR + AI step.
Proof · POPIA impact assessment for a real workflow, expert-reviewed.
1 lesson
Defend
Advanced
Run an AI incident response
Detect, contain, communicate, learn — when the model gets it wrong.
Proof · Tabletop exercise transcript with rubric-scored decisions.
1 lesson
Defend
Foundational
Spot bias in real HR copy
Pattern-match bias in JDs, performance reviews, and policy text on sight.
Proof · ≥ 80% accuracy across the Bias Spotter rounds.