AI glossary for business owners

Term	Meaning
AI agent	An AI system that can take steps toward a goal, such as looking up data, filling a form or sending a message, within permissions you set.
Context window	How much text a model can consider at once, including your documents and the conversation.
Embedding	A numerical representation of text that lets software find passages with similar meaning.
Evaluation (evals)	Systematic testing of AI outputs against expected answers or quality criteria.
Fine-tuning	Further training a model on your examples to change its style or behaviour.
Grounding	Making the AI answer from specific source material rather than general knowledge.
Guardrails	Rules and checks that keep AI outputs safe, accurate and on-topic.
Hallucination	A confident answer that isn't supported by facts or sources.
Human-in-the-loop	A person reviews or approves AI output before it takes effect.
LLM	Large language model: AI trained on large amounts of text to understand and generate language.
Prompt	The instructions and input given to an AI model.
Prompt injection	An attempt to make an AI ignore its instructions through crafted input.
RAG	Retrieval-augmented generation: searching your documents and giving relevant passages to the model before it answers.
Token	A chunk of text that models process; usage is often priced per token.
Vector database	A database that stores embeddings so similar content can be found quickly.

More in Reference