A

‍

Artificial Intelligence (AI)

Artificial Intelligence is the broad field of technology focused on building systems that can perform tasks that normally require human intelligence, such as understanding language, recognising patterns, or making decisions.

‍

Agent

An agent is a software-based entity that can observe information, decide what to do next, and take action to achieve a goal.

‍

AI Agent

An AI-powered agent that uses models such as large language models to reason, plan, and perform tasks, often by using tools or accessing data.

‍

Agentic AI

Agentic AI refers to AI systems that are designed to act, not just respond. These systems can plan steps, use tools, remember context, and complete tasks semi-autonomously within defined limits.

‍

Autonomous Agent

An autonomous agent is an AI agent that can operate with minimal human input while still following rules, permissions, and safety constraints.

‍

AI Adoption

The process of integrating AI into daily workflows, tools, and decision-making across an organisation.

‍

AI Readiness

A measure of how prepared an organisation is to successfully use AI, including data quality, skills, systems, and governance.

‍

AI Maturity

The level at which an organisation has progressed in its AI usage, from experimentation to structured, scalable deployment.

‍

C

‍

CAIO (Chief AI Officer)

A leadership role responsible for overseeing AI strategy, ethics, governance, and ensuring AI is used responsibly and effectively.

‍

CTO (Chief Technology Officer)

The executive responsible for an organisation’s technology infrastructure and technical decisions, including AI platforms and integrations.

‍

ChatGPT

A conversational AI product built on large language models that allows users to interact with AI using natural language.

‍

Context Window

The maximum amount of information an AI model can consider at one time when generating a response.

‍

Context Injection

The practice of supplying additional background information, instructions, or data to an AI model to improve relevance and accuracy.

‍

D

‍

Deep Learning

A subset of machine learning that uses neural networks with many layers to recognise complex patterns in data such as text, images, or speech.

‍

Digital Workforce

A working model where humans and AI agents collaborate, with AI handling repetitive or data-heavy tasks and humans focusing on judgment and creativity.

‍

Document Grounding

A technique where AI responses are constrained to specific documents or sources, reducing hallucinations and improving trustworthiness.

‍

E

‍

Embeddings

Numerical representations of text or data that capture meaning, allowing AI systems to compare, search, and retrieve information based on similarity.

‍

EU AI Act

The European Union’s legal framework regulating AI systems, categorising them by risk and setting rules for transparency, safety, and accountability.

‍

Explainability

The ability to understand and describe how an AI system arrives at its outputs or decisions.

‍

Event-Driven Automation

Automation that is triggered by specific events, such as a new email, form submission, or system update.

‍

Exception Handling

The way AI or automation systems detect, manage, and escalate situations that fall outside normal rules or expected behaviour.

‍

F

‍

Fine-tuning

The process of adapting a pre-trained AI model to a specific domain or task using additional, targeted data.

‍

Foundation Model

A large, general-purpose AI model trained on broad datasets and adapted for many different tasks.

‍

Function Calling

A capability that allows AI models to trigger predefined software functions or actions in external systems.

‍

G

‍

GPT (Generative Pre-trained Transformer)

A type of AI model architecture designed to understand and generate natural language based on patterns learned during training.

‍

Gemini

A family of AI models developed by Google, designed for multimodal reasoning across text, images, and other data types.

‍

H

‍

Hallucination

When an AI system produces output that sounds correct but is factually wrong or unsupported by data.

‍

Human-in-the-loop

A design approach where humans maintain oversight, approval, or control over AI actions, especially in sensitive or high-impact situations.

‍

Human–AI Collaboration

A working model where humans and AI systems complement each other, combining speed and automation with human judgment and empathy.

‍

I

‍

Inference

The process by which a trained AI model generates outputs based on new input data.

‍

Intelligent Automation

Automation that combines traditional workflows with AI capabilities such as reasoning, language understanding, or decision support.

‍

K

‍

Knowledge Base

A structured collection of documents or information used by AI systems to retrieve accurate and relevant answers.

‍

L

‍

Large Language Model (LLM)

A type of AI model trained on very large amounts of text to understand and generate human language.

‍

LangChain

A framework used to build AI applications by connecting language models with tools, memory, and workflows.

‍

M

‍

Machine Learning (ML)

A branch of AI where systems learn patterns from data and improve over time without being explicitly programmed.

‍

Model

A trained AI system that performs a specific task, such as text generation or classification.

‍

Model Bias

Systematic errors in AI outputs caused by imbalances or issues in training data.

‍

Model Alignment

The practice of ensuring AI systems behave in line with human values, rules, and intended goals.

‍

Multimodal Model

An AI model that can process multiple types of input, such as text, images, audio, or video.

‍

Memory (Short-term / Long-term)

Mechanisms that allow AI agents to retain information temporarily (short-term) or persist knowledge over time (long-term).

‍

N

‍

n8n

A workflow automation platform that allows users to connect applications, APIs, and AI models using visual workflows.

‍

NotebookLM

An AI tool designed to analyse, summarise, and reason over user-provided documents and notes.

‍

O

‍

OpenAI

An AI research and deployment organisation known for developing GPT models and ChatGPT.

‍

OpenAI API

A developer interface that allows applications to integrate AI capabilities such as language generation and reasoning.

‍

Orchestration

The coordination of multiple AI agents, tools, and workflows to execute complex tasks in the correct order.

‍

P

‍

Prompt

The input or instruction provided to an AI model that guides its response.

‍

Prompt Engineering

The practice of designing prompts to improve the quality, accuracy, and consistency of AI outputs.

‍

Prompt Chaining

Linking multiple prompts together so the output of one step becomes the input for the next.

‍

Planning

The ability of an AI system to break a goal into smaller steps and decide the order of actions.

‍

Proof of Concept (POC)

A small-scale implementation used to test whether an AI idea works before wider deployment.

‍

R

‍

Retrieval-Augmented Generation (RAG)

An AI approach where a language model retrieves information from external sources before generating a response.

‍

Reflection Loop

A mechanism where an AI system reviews and improves its own outputs to reduce errors.

‍

Responsible AI

The practice of developing and using AI in ways that are ethical, transparent, safe, and accountable.

‍

Robotic Process Automation (RPA)

Technology that automates repetitive, rule-based tasks by mimicking human interactions with software.

‍

S

‍

Semantic Search

Search technology that focuses on meaning rather than exact keyword matches, often powered by embeddings.

‍

System Prompt

A high-level instruction that defines an AI model’s role, behaviour, and constraints.

‍

T

‍

Token

A small unit of text processed by an AI model, such as part of a word or symbol.

‍

Training

The process of teaching an AI model patterns by exposing it to large datasets.

‍

Temperature

A setting that controls how predictable or creative an AI model’s responses are.

‍

Top-p (Nucleus Sampling)

A technique that limits AI responses to the most likely options, balancing creativity and reliability.

‍

Tool Calling

A capability that allows AI agents to use external tools, APIs, or software systems.

‍

V

‍

Vector Database

A specialised database designed to store and search embeddings efficiently.

‍

Vector Embedding

A numerical representation of data that captures meaning and enables similarity comparison.

‍

W

‍

Workflow Automation

The automation of multi-step processes across systems and tools, often combining rules, triggers, and AI.

‍

This glossary is designed to evolve as AI technology, regulation, and best practices continue to develop.

______