AI glossary

A short explanation, followed by a path to the full guide. Start with the term you ran into; you don’t need to learn them all at once.

Agent

A system that uses a model to choose some actions and next steps toward a goal.

Read the full guide

API

A defined interface through which software requests data or an action from other software.

Read the full guide

Automation

A process that runs steps when a trigger or schedule occurs. It may use no AI at all.

Read the full guide

Benchmark

A shared test for comparison. Its score may not predict performance on your actual task.

Read the full guide

Connector

An integration that lets an application access another service within granted permissions.

Read the full guide

Context window

The bounded amount of information available to a model in a request.

Read the full guide

Embedding

A numerical representation used to compare content or other inputs.

Read the full guide

Evaluation / eval

A test of whether a system meets defined quality and behavior criteria.

Read the full guide

Fine-tuning

Additional training that changes a model’s learned parameters using selected examples.

Read the full guide

Foundation model

A broadly trained model that can be adapted or used for different tasks.

Read the full guide

Generative AI

Systems that produce content such as text, images, audio, or code.

Read the full guide

Governance

The people, rules, evidence, and decisions that make AI use accountable.

Read the full guide

Guardrail

A control intended to constrain behavior. It needs testing and can fail.

Read the full guide

Hallucination

An output that sounds plausible but contains unsupported or incorrect information.

Read the full guide

Human in the loop

A person who can review or intervene at a meaningful point in a system.

Read the full guide

Inference

Using a trained model to produce an output from new input.

Read the full guide

LLM

A large language model: a model trained to work with language, often alongside other modalities.

Read the full guide

MCP

Model Context Protocol: a standard for connecting AI applications to tools and contextual data.

Read the full guide

Memory

Information an application preserves and makes available later. It is a designed feature.

Read the full guide

Multimodal

Working with more than one type of information, such as text, images, and audio.

Read the full guide

Orchestration

Coordinating steps, models, tools, or agents, including state and failure handling.

Read the full guide

Plugin

A packaged extension that adds capabilities to an application; its contents vary by platform.

Read the full guide

Prompt injection

Untrusted input that tries to redirect an AI system’s behavior or authority.

Read the full guide

RAG

Retrieval augmented generation: finding relevant material and using it as context for generation.

Read the full guide

Skill

Reusable task instructions, sometimes bundled with examples or code, in a supporting application.

Read the full guide

Structured output

A response constrained to a format or schema. Valid structure does not establish truth.

Read the full guide

Token

A unit used to represent input and output for a model. Tokenization depends on the model.

Read the full guide

Tool calling

A model requesting an available function; the surrounding application executes and controls it.

Read the full guide

Finding nearby numerical representations, often to retrieve semantically related material.

Read the full guide

Workflow

An organized process with inputs, steps, outputs, and handling for exceptions.

Read the full guide

Share this page