Six free lessons · About 70 minutes

Testing and quality control in the age of AI

AI can make a first draft fast. This series is about finding out whether the result actually works, whether its claims have support, and what should happen before anyone relies on it.

Start part one Beginner-friendly; code example optional

The idea

Trust needs checkpoints.

A polished website, a busy campaign, and a confident AI review can all hide a broken result. The answer is not to stop using AI. It is to give the work a few checks that are clear enough for a teammate to repeat.

We begin with the whole visitor or user journey, expand the cases a team can afford to test, add a focused reviewing agent, and place checks at risky handoffs. Then we show how experienced quality teams can use AI for more reach and how to set stable rules for accepting variable output. Every lesson has a realistic teaching scenario, a cartoon, and a small assignment.

Cartoon teammate checks a web page while a friendly AI helper offers a tall stack of draft screens.

Part 1 of 612 minutes

How to Test Websites, Campaigns, and AI-Built Software

A business-first way to check whether a website, marketing campaign, or AI-built tool actually delivers its promised result. Playwright is one optional way to automate the checks.

Read part 1

Cartoon tester and AI helper sort many web-page cards, flagging one for review.

Part 2 of 612 minutes

How AI Expands What Your Team Can Test

See how AI makes a first-pass review of many pages and claims practical for a small team—and where human sampling is still essential.

Read part 2

Cartoon AI detective examines a draft while a human reviewer points to a source document.

Part 3 of 612 minutes

How to Use AI Reviewers and Check Sources

Give an AI reviewer a narrow assignment, ask for the exact source behind important claims, and keep a person responsible for the final call.

Read part 3

Cartoon conveyor carries draft pages through inspection gates while an incorrect page is diverted for review.

Part 4 of 611 minutes

Build Verification Checkpoints for AI Work

Place small checks where an AI draft becomes a decision, a saved record, or a customer-facing action.

Read part 4

Cartoon quality lead directs two friendly AI helpers that sort test cards and inspect a suspicious draft.

Part 5 of 613 minutes

Help Your Quality-Assurance Team Do More With AI

Keep your experienced reviewers in charge of risk and release decisions while AI helps draft cases, prepare data, sort failures, and maintain repeatable checks.

Read part 5

Cartoon teammate feeds an AI draft through a transparent machine of checks while a mischievous bad draft is stopped.

Part 6 of 612 minutes

Turn AI Drafts Into Repeatable, Testable Work

The model may stay creative; the rules for accepting its work can be stable, measurable, and owned by people.

Read part 6

What you can use right away

Leave with a test brief, a broader coverage plan, a source-review prompt, and a stop rule for your next release. A passing check alone does not prove every AI output is true or every customer journey works. The habit is to name the risk, test the outcome, keep the evidence, and know when a person must decide.