AI dictionary

The artificial intelligence terms that come up when you run a business, defined in plain English and sourced.

A

Agent

An AI tool that takes a series of actions on its own to complete a goal you set.

C

Context Window

The amount of text an AI tool can hold in mind at once, including its own reply.

F

Fine-Tuning

Training a general AI model further on your own examples so it excels at a specific job.

H

Hallucination

When an AI tool gives a confident, plausible answer that is actually wrong or made up.

I

Inference

The moment an AI model is used to turn an input into an answer, drawing on its training.

L

Large Language Model

LLM

An AI trained on huge amounts of text to understand and write language.

M

Model Evaluation

Testing an AI model to measure how well it performs before you rely on it.

P

Prompt Engineering

Writing clear, well-structured instructions to get a reliable answer from an AI tool.

R

Retrieval Augmented Generation

RAG

Letting an AI tool look up trusted documents before it answers, to ground its reply.

T

Training Data

The examples an AI model learns from, which shape what it can and cannot do.

The guides are free. The checks are the product.

Run the same checks on a client site and get the top findings for search, conversion and accessibility in minutes.

Get a free website auditSee pricing