Question 1

Task: "Plan tests for this login page and write the 3 most important ones." Your approach?

Accepted Answer

First I talk through a 1-2 minute plan: positive (successful login), negative (wrong password, empty fields, unknown user), security (lockout, injection), UX (error messages). Then justify priorities: (1) successful login — the critical path; (2) wrong credentials + correct error — the most common real case; (3) validation (empty fields). In code: proper locators, web-first assertions, users from a data factory. With time left I offer a refactor (extract to POM). Process: plan → priorities → code → refactor.

Question 2

JS exercise: "From an API order array, find the top 3 spenders." Your approach?

Accepted Answer

Steps aloud: (1) understand the data — ask: can one customer have several orders? (yes → aggregation needed); (2) plan: reduce to per-customer sums → object to array → sort desc → slice(0, 3); (3) write the code, flagging edges: empty array, ties, do cancelled orders count? Sample: Object.entries(orders.reduce((acc, o) => ({...acc, [o.customerId]: (acc[o.customerId] ?? 0) + o.total}), {})).sort((a, b) => b[1] - a[1]).slice(0, 3). Asking edge questions is valued as much as the code.

Question 3

Debug task: the test fails with "element not found" but the element is on the page. Investigate aloud?

Accepted Answer

I walk hypotheses: (1) timing — is the element rendered later? does the assertion auto-retry, what's awaited before the action? (2) locator — inside an iframe (needs frameLocator)? shadow DOM? dynamic id? (3) strict mode — maybe multiple matches with a different error message; (4) context — right page/tab? navigation finished? (5) open the trace — is the element truly in the failure snapshot. I state each hypothesis with its verification method — demonstrating systematic debugging matters more than the answer itself.

Live coding tasks