AI & software terms, in plain English

AI Agent

An AI agent is a system that uses a language model to plan and take actions — calling tools, querying data and...

Chain-of-Thought

Chain-of-thought is a prompting technique that asks a model to reason step by step before giving a final answe...

Context Window

A context window is the maximum amount of text (measured in tokens) a language model can consider at once — bo...

Embeddings

An embedding is a numerical vector that represents the meaning of text, an image or other data, so that semant...

Fine-tuning

Fine-tuning adapts a pre-trained model to a specific task or domain by continuing training on a smaller, targe...

Foundation Model

A foundation model is a large model trained on broad data at scale that can be adapted to many downstream task...

Guardrails

Guardrails are the controls that keep an AI system safe and on-topic — input/output validation, content filter...

Hallucination

A hallucination is when an AI model generates output that is fluent and confident but factually wrong or unsup...

Inference

Inference is the process of running a trained model to produce an output — for example generating a response o...

Large Language Model (LLM)

A large language model is an AI system trained on vast amounts of text to understand and generate human-like l...

Model Context Protocol (MCP)

The Model Context Protocol is an open standard for connecting AI assistants to external tools and data sources...

Model Distillation

Distillation trains a smaller, cheaper "student" model to mimic a larger "teacher" model, keeping most of the...

Multimodal AI

Multimodal AI refers to models that understand and generate more than one type of data — for example text, ima...

Prompt Engineering

Prompt engineering is the practice of designing the instructions and context given to a language model to get...

Retrieval-Augmented Generation (RAG)

RAG is a technique that grounds a language model's answers in your own data by retrieving relevant documents a...

RLHF

Reinforcement Learning from Human Feedback (RLHF) is a training method that aligns a model with human preferen...

Semantic Search

Semantic search finds results by meaning rather than exact keywords, using embeddings to match a query to conc...

Token

A token is the unit of text a language model reads and generates — roughly a word or word-fragment. Models pri...

API Gateway

An API gateway is a single entry point that sits in front of backend services, handling routing, authenticatio...

Event-Driven Architecture

Event-driven architecture lets components communicate by producing and reacting to events rather than calling...

Microservices

Microservices is an architecture that builds an application as a set of small, independently deployable servic...

Multi-Tenancy

Multi-tenancy is an architecture where a single application instance serves many customers (tenants) while kee...

CI/CD

CI/CD (Continuous Integration / Continuous Delivery) is the practice of automatically building, testing and de...

Infrastructure as Code

Infrastructure as Code (IaC) manages servers, networks and cloud resources through version-controlled configur...

MLOps

MLOps is the set of practices for deploying, monitoring and maintaining machine-learning models reliably in pr...

Observability

Observability is the ability to understand what is happening inside a system from its outputs — logs, metrics...

Quantization

Quantization shrinks a model by storing its weights at lower numerical precision (for example 8-bit instead of...

Technical Debt

Technical debt is the future cost of shortcuts taken to ship faster today — code or design that will need rewo...

Data Lake

A data lake stores raw data of any type — structured, semi-structured and unstructured — at low cost and large...

Data Pipeline

A data pipeline is an automated flow that moves and processes data from sources to destinations — ingesting, c...

Data Warehouse

A data warehouse is a system optimised for analytical queries over large volumes of structured data, consolida...

ETL / ELT

ETL (Extract, Transform, Load) and ELT are patterns for moving data from source systems into a warehouse or la...

Feature Store

A feature store is a system that manages the input variables ("features") used by machine-learning models, ens...

Knowledge Graph

A knowledge graph represents information as entities and the relationships between them, letting systems answe...

Vector Database

A vector database stores data as high-dimensional numerical embeddings and retrieves items by semantic similar...

Kubernetes

Kubernetes is an open-source platform that automates deploying, scaling and managing containerised application...

Serverless

Serverless is a cloud model where you run code without managing servers — the provider allocates resources on...

Legacy Modernization

Legacy modernization is the process of upgrading outdated software — re-platforming, refactoring or rebuilding...

Minimum Viable Product (MVP)

A Minimum Viable Product is the smallest version of a product that delivers real value and validates a hypothe...

Total Cost of Ownership (TCO)

Total Cost of Ownership is the full lifetime cost of a system — not just the upfront price, but licensing, hos...

Prompt Injection

Prompt injection is an attack where malicious instructions hidden in content (a web page, document or email) t...

Single Sign-On (SSO)

Single Sign-On lets users access multiple applications with one set of credentials, typically via an identity...

SOC 2

SOC 2 is a widely recognised audit standard that assesses how well a service organisation protects customer da...

Zero Trust

Zero Trust is a security model that trusts no user or device by default — every request is authenticated, auth...