Resources

The AI glossary & resources

Plain-English definitions of the AI terms that matter — plus tools to go deeper.

AI glossary

Every AI term, in plain English.

The vocabulary behind modern AI — no jargon, no hand-waving. Start typing to jump straight to a term.

AI Agent

Software that uses an AI model to pursue a goal on its own — it decides what to do, takes actions across your tools and APIs, and reacts to the results. Not just answering, but doing.

Agentic AI

AI that plans and acts over multiple steps toward a goal, choosing its own actions and adjusting as it goes — rather than responding to a single prompt in isolation.

LLM (Large Language Model)

A model trained on huge amounts of text to understand and generate human language. It powers chatbots, copilots and most modern AI features by predicting the most likely next words.

RAG (Retrieval-Augmented Generation)

A technique that fetches relevant information from your own data and feeds it to the model at answer time — so responses are grounded in real facts instead of the model's memory alone.

Fine-tuning

Further training a pre-trained model on your own examples so it adapts to a specific tone, task or domain — teaching an existing model new habits rather than building one from scratch.

Embeddings

Numeric representations of text or images that capture meaning, letting a computer measure how similar two pieces of content are. They're the backbone of AI search and RAG.

Vector Database

A database built to store embeddings and find the most similar ones fast. It's what lets an AI app retrieve the right context from thousands of documents in milliseconds.

Prompt Engineering

The craft of writing clear instructions, context and examples for an AI model so it returns accurate, reliable, on-task results — the difference between a vague answer and a useful one.

Hallucination

When an AI model produces confident but false or made-up information. Techniques like RAG, guardrails and citations are used to catch and reduce it.

Guardrails

Rules and checks around an AI system that keep it safe and on-policy — blocking harmful, off-topic or risky outputs and actions before they reach the user or your systems.

MCP (Model Context Protocol)

An open standard that lets AI models connect to external tools, data and systems through one common interface — so an agent can plug into your apps without custom glue for each one.

AI Digital Employee

An AI agent scoped to own a whole role — like support, ops or research — working 24/7 inside your systems with guardrails, instead of handling just a single one-off task.

Inference

Running a trained model to get an output — actually using the AI to answer a question or make a prediction, as opposed to training it. Every response the model gives is an inference.

Token

The small chunk of text — roughly a few characters or part of a word — that models read and generate. AI usage and pricing are usually measured in tokens.

Context Window

The maximum amount of text, measured in tokens, a model can consider at once — its short-term memory. Anything beyond it has to be summarised, dropped or retrieved on demand.

Multimodal AI

AI that can understand or generate more than one type of content — text, images, audio and video — rather than being limited to text alone.

Human-in-the-loop

A design where a person reviews, approves or corrects the AI at key steps — keeping humans firmly in control of sensitive or high-stakes decisions.

Audit Trail

A complete, tamper-evident record of what an AI system did and why — every action, input and approval — so decisions can be reviewed, explained and trusted.

Go deeper

Tools to take it further.

Beyond the definitions — see the work, price your build, read the thinking, or just talk to us.

Let's build

From glossary to shipped product.

Now you know the vocabulary — let's put it to work. Tell us what you want to build and we'll come back with a clear, practical plan and a call to map the fastest path.

Book a call → See live demos →

What we build