How to use this AI glossary
This glossary translates terms that appear in product, engineering, and artificial-intelligence operations. Each definition starts with a direct meaning and adds the context needed for a decision: when the concept is useful, which limits matter, and what evidence helps evaluate an implementation.
Use the page to align language across product, data, security, and business teams. Treat a term as a working model rather than a promise, and look for the relationship between capability, cost, risk, and outcome. RAG, tool calling, evaluation, and inference each carry different implications for architecture, user experience, and governance.
Entries are updated when new sources change an explanation or when an operational example helps prevent a common misunderstanding. For applied guidance, combine this glossary with the AI guides and the topic maps.
When a term affects a roadmap or a control, use its entry as a shared starting point, then document the local definition, assumptions, and evidence your team will use in practice.