New: the QA x AI radar, the glossary and 8 ready-to-copy QA prompts.See the radar →

define('glossary')

QA x AI glossary

QA and AI terms explained simply: flaky test, test oracle, LLM-as-a-judge, prompt injection... A short definition, then what you need to put it into practice.

C

  • Contract testingCI/CD and APIs

    Checking that an API and its clients respect the same contract, without a full environment.

E

  • End-to-end test (E2E)Testing

    A test that reproduces a complete user journey, from the interface to the database.

  • Exploratory testingMethods

    Learning, designing and running tests at the same time, without a fixed script.

F

  • Flaky testTesting

    A test that passes or fails without the code having changed.

G

  • Golden datasetAI and LLMs

    A set of test cases whose correct answer has been validated by a human.

H

  • HallucinationAI and LLMs

    An AI answer that sounds plausible but is wrong or made up.

L

  • LLM-as-a-judgeAI and LLMs

    Using a language model to evaluate another model's answers.

P

  • Prompt injectionAI and LLMs

    An attack that hijacks an AI by slipping instructions into its data.

R

  • Red teamingAI and LLMs

    Deliberately attacking a system to find its weaknesses before others do.

  • Regression testingTesting

    Checking that a change hasn't broken what already worked.

S

T

  • Test impact analysisCI/CD and APIs

    Only running the tests affected by a code change.

  • Test oracleTesting

    The mechanism that decides whether the actual result is the right one.

  • Test pyramidMethods

    Lots of unit tests, fewer integration tests, a few end-to-end tests.

A term is missing?

The glossary grows with your questions. Suggest a word and it will be defined in the next update.

Suggest a term