This product is the assessment

Your software. Your public repository. Your decisions.

AI assistance is encouraged and disclosed.

Jared reviews your product, development history and GitHub practice.

No separate exam. No surprise coding challenge.

Eight minutes to tell the story

1 minute: audience, problem and scope.

3 minutes: working journey, saved result, error and recovery.

2 minutes: GitHub review story and meaningful regression.

1 minute: AI correction, cost and business decision.

1 minute: handoff, limits and next improvement.

Show something people can use

Clear visual hierarchy and consistent navigation.

Readable mobile layout, keyboard access and helpful states.

Synthetic data with a complete end-to-end journey.

Polish supports usefulness; screenshots alone do not prove it.

Open the evidence

Issue → topic branch → focused PR → human feedback → fix.

A project defect or labeled approved control: fail before / pass after.

Design iterations, boundary checks and recovery.

Link exact revisions; state what was not tested.

Explain your AI workflow

Which model, client and host did you use?

What did AI get wrong, and how did you catch it?

How did smaller tasks and durable context help?

What did it cost? Which measurements are missing?

Four minutes with Jared

Discuss choices already represented in your product.

Explain one tradeoff, correction or remaining limitation.

Use your notes and evidence; this is not a memory quiz.

Ask for advice on the next useful improvement.

A reliable handoff

Reviewed release, setup/run/check/reset instructions.

Safe sample data, source acknowledgments and known limits.

Local demo is valid; paid hosting is not required.

Keep a backup demo and close private tabs and notifications.

Course completion ≠ skills certification

Completion certificate: finished learning activities and presentation.

Labalicious Certified AIQAA: demonstrated product quality,
development over time and learned GitHub/QA skills.

Jared records and authorizes each award separately.

Show the evolution, not a commit counter

Early plan → intermediate build → test/fix → reviewed release.

Explain a diff, review response and meaningful verification.

Use real PRs, milestone evidence and instructor checkpoints.

A polished final upload alone cannot establish learned practice.

Review the evidence, not an impression

Frozen agreement + release revision + all 21 criteria.

Demonstrated: observed conditions hold.

Not yet: a condition failed. Unverified: evidence is unresolved.

All required criteria demonstrated; Jared authorizes the award.

Not yet certified? Keep improving.

Completed coursework still receives its completion certificate.

Certification gaps get clear feedback and another review.

No separate exam, points average or mandatory premium model.

Private decisions and issuance; no automatic certification.

Skills that keep growing

Exploratory QA · automation · integration reliability

Accessibility · visual quality · AI evaluation

Outcome → context → constraints → evidence → review

Models change. Keep improving the workflow.