Artificial intelligence glossary

This artificial intelligence glossary defines, in one sentence, the key terms you will meet in our courses: machine learning, generative AI, LLMs, agents, RAG and automation. Each definition is self-contained and links, where relevant, to the related course.

Artificial intelligence (AI)

Artificial intelligence refers to computer systems able to perform tasks that usually require human intelligence: understanding language, recognising images, deciding or generating content.

Machine learning

Machine learning is a branch of AI where a model learns patterns from data rather than being programmed with explicit rules.

Deep learning

Deep learning is machine learning based on neural networks with many layers, especially effective on images, sound and language.

Neural network

A neural network is a brain-inspired model made of layers of ‘neurons’ that transform inputs into outputs through learning.

Generative AI

Generative AI creates new content (text, image, code, audio) from a prompt, using models trained on vast data. See our generative AI course.

LLM (large language model)

An LLM is an AI model trained on huge text corpora to understand and generate natural language (Claude, GPT…). See our Claude course.

Prompt

A prompt is the natural-language instruction given to a generative AI to obtain a response.

Prompt engineering

Prompt engineering is the craft of writing effective instructions to get better results from a generative AI.

Token

A token is the basic unit (word or word fragment) an LLM processes; input and output length is measured in tokens.

Hallucination

A hallucination is a plausible but false or made-up AI response; checking the output remains essential.

RAG (retrieval-augmented generation)

RAG enriches an LLM’s answers with documents retrieved from a database, to ground responses in your data.

Fine-tuning

Fine-tuning adjusts a pre-trained model on specific data to adapt it to a domain or a task.

AI agent

An AI agent is a system that uses an LLM to reason and chain actions (calling tools, querying data) to achieve a goal. See our AI agents on Azure course.

Copilot

A copilot is an AI assistant embedded in software (e.g. Microsoft 365 Copilot) that helps the user with their tasks. See our Microsoft 365 Copilot course.

Chatbot

A chatbot is a conversational program that talks in natural language; the newest ones rely on LLMs.

NLP (natural language processing)

NLP covers the techniques that let a machine understand and produce human language.

Computer vision

Computer vision lets a machine analyse and interpret images and videos.

Embedding

An embedding is a numerical representation (vector) of a text or image that captures its meaning and enables similarity search.

Foundation model

A foundation model is a large, versatile pre-trained model reusable for many tasks (often via prompting or fine-tuning).

MLOps

MLOps applies DevOps principles to machine learning: industrialising model training course, deployment and monitoring.

Bias / AI ethics

A bias is a distortion in a model’s results inherited from training course data; AI ethics aims to limit these biases and harmful uses.

Automation (workflow)

Automation orchestrates repetitive tasks across applications; tools like n8n now embed AI steps. See our n8n course.