Glossary

The vocabulary of AI testing and the AI-driven development lifecycle - defined, sourced, and honest about what the industry has not agreed on.

  • AI Testing and AI-DLC: A Glossary

    The vocabulary of AI testing is moving faster than anyone's ability to standardize it. Terms get coined in a blog post, picked up by vendors, and are load-bearing in a procurement document six months later — often meaning three different things to the ...

    Read more

  • The AI-Driven Development Lifecycle: Terms

    The vocabulary of building software with AI agents — the lifecycle framings, the practices, the artifacts, and where verification sits inside them. ABCDEFGHIJKLMNOPQRSTUVWXYZ A Adversarial review Review by a separate agent with a fresh context, which se...

    Read more

  • Agentic Systems: Terms

    The vocabulary of agents themselves — how they are built, what they remember, how they connect to your systems, and the named ways they fail. ABCDEFGHIJKLMNOPQRSTUVWXYZ A Agentic system An umbrella term covering both workflows, where models and tools ar...

    Read more

  • Evaluating and Testing AI: Terms

    The vocabulary of proving an AI system works — evals, judges, groundedness, red teaming, and the statistics you need when a test case is a distribution rather than a pass or a fail. ABCDEFGHIJKLMNOPQRSTUVWXYZ A A/B testing of models also: champion/chall...

    Read more

  • Quality Engineering and AI Assurance: Terms

    Classic quality-engineering vocabulary restated precisely, because AI vendors blur it — plus the governance, standards and regulatory terms that now touch testing. ABCDEFGHIJKLMNOPQRSTUVWXYZ A Accuracy, robustness and cybersecurity Article 15 requires h...

    Read more