QASubsection
Types of testing
Three tests that people put on one list and that are not comparable. Smoke is a gate: it decides whether a build is worth anyone else’s time. Regression is memory: it remembers what already broke once. End-to-end is a level: it is the only place the assembled system can be proved to work. The first two are reasons to run tests and can be written at any level, which is the distinction most interview answers miss.
What is inside
Every page in this group, with what each one covers.
- Smoke TestingThe selection rule, the time budget as a mechanism, and the difference between smoke, sanity and confirmation.GateUnder 5 min
- Regression TestingWhat earns a permanent case, the stricter exit rule, five selection strategies, and quarantine over delete.MemorySelection
- E2E TestingWhat only the assembled system can prove, why the tier is small rather than absent, and choosing journeys not features.JourneysPyramid
Where to start
- Building a suite from nothingSmoke testing. A small suite that proves the system is up beats a large one nobody trusts.
- The suite is too slow to runRegression testing, and the selection section in particular — parallelise first, tier second.
- Your E2E tests are flakyEnd-to-end testing — the five sources of variation, and what should have been caught two levels down.
Elsewhere in the section
- QAQuality assurance is the practice of deciding what to check, running those checks often enough to be worth having, and knowing what changes when the thing under test stops giving the same answer twice. This section covers 9 topics in three groups: the types of testing everyone half-remembers, automation with Playwright, Cypress and Selenium, and testing products built on language models — where an assertion becomes a score rather than an equality.
- AutomationThree browser tools, each explained through the one architectural fact that produces everything else about it. Playwright controls a browser from outside over a single connection; Cypress runs inside the tab with your application; Selenium is a published W3C standard with libraries attached. Read the architecture and the feature lists explain themselves — including the parts each tool cannot do. What belongs in a browser test at all is on the E2E page, not here.
- AI TestingWhen the same input stops producing the same output, an equality assertion becomes meaningless and a single run stops being evidence. Three topics: what that changes for the person writing the cases, how to judge one answer without spending money to replace a fact with an opinion, and how to tell a real regression from sampling noise. Running an LLM feature in production is LLMOps, in Development — these three own testing it.
Found this useful?
Share it with someone who is working on the same problem.