AI glossary

AI terms explained clearly.

Concise definitions, technical background and links to related guides, topics and services.

A

Agentic AI

Agentic AI describes AI systems that independently plan multi-step tasks, call tools, and evaluate intermediate results, instead of answering a single request.

View term

AI Gateway

An AI gateway is a central layer between applications and AI models that bundles access, logs requests, enforces security rules, and can route between different providers.

View term

C

Context Window

The context window is the maximum amount of text, measured in tokens, that a language model can process at once in a single request.

View term

E

Embedding

An embedding represents text, images or other data as a numeric vector. Semantically similar data receives similar vectors.

View term

F

Fine-Tuning

Fine-tuning further trains an already trained model on additional, specific data to specialise it for a particular task or style.

View term

G

Guardrails

Guardrails are technical and organisational safeguards that define what an AI system may do, which data it may see and when human review or approval is required.

View term

H

Hallucination

A hallucination is a plausibly phrased but factually incorrect or fabricated output from an AI model.

View term

I

Inference

Inference is the process where an already trained AI model processes an input and produces an output, as opposed to training, where the model itself is adjusted.

View term

L

Large Language Model (LLM)

A large language model is an AI model trained on large amounts of text that processes language patterns to generate, summarise and edit text.

View term

M

MCP (Model Context Protocol)

MCP is an open standard that lets AI models access external tools, data, and systems in a uniform way, instead of building a custom integration for every connection.

View term

P

Prompt Engineering

Prompt engineering is the systematic design of instructions given to a language model, aimed at getting more reliable, precise, or better structured answers.

View term

R

RAG (Retrieval-Augmented Generation)

RAG combines a language model with a search over your own documents, so answers are grounded in current, company-specific knowledge instead of training data alone.

View term

T

Token

A token is a processing unit for text, often a word, word fragment or punctuation mark. Tokens affect cost, speed and context size.

View term

Tool Use (Function Calling)

Tool use is a language model's ability to call external functions in a structured way while forming an answer, for example a search, a calculation, or a database query.

View term

V

Vector Database

A vector database stores embeddings and retrieves entries according to their mathematically determined similarity.

View term

What would you like to assess or build?

Briefly describe the situation, your objective and the systems involved. We will respond with a technical assessment and a practical next step.

Send inquiry