Agentic AI Glossary, Plain-English Definitions

This glossary is maintained by Eduk8agentic as part of our mission to make agentic AI accessible to business people who don’t code.

Core Concepts

Agentic AI

Artificial intelligence that can take autonomous actions, make decisions, and complete multi-step tasks on your behalf, without requiring constant human supervision. Unlike a chatbot that responds to a single question, an agentic AI system can plan a sequence of steps, use tools, and work towards a goal over an extended period.

Example: An agentic AI system asked to "research our top three competitors and produce a pricing comparison" will independently search the web, read competitor pages, extract pricing data, and format a comparison table, all without further instruction.

AI Agent

A software system powered by a large language model that can perceive its environment, make decisions, and take actions to achieve a specified goal. An AI agent combines reasoning capabilities with access to tools (such as web search, file reading, or API calls) to complete complex tasks autonomously.

Example: A legal AI agent might be given the goal of reviewing a contract. It reads the document, identifies unusual clauses, searches for relevant legal precedents, and produces a risk summary, all as a single autonomous workflow.

Large Language Model (LLM)

A type of AI system trained on vast amounts of text data that can understand and generate human language. LLMs are the foundation of modern AI assistants and agents. Examples include Claude (by Anthropic), GPT-4 (by OpenAI), and Gemini (by Google).

Example: When you ask Claude to summarise a document or draft an email, you are interacting with a large language model.

Workflow

A defined sequence of steps that an AI agent follows to complete a task. Workflows break complex goals into manageable steps, making AI systems more reliable and their outputs more predictable.

Example: A content research workflow might involve: (1) search for recent articles on the topic, (2) extract key points, (3) identify common themes, (4) produce a structured summary with citations.

Tool Use

The ability of an AI agent to interact with external tools and services, such as web browsers, spreadsheets, databases, or APIs, to gather information and take actions beyond just generating text.

Example: An AI agent with tool use can search Google for competitor pricing, read a PDF contract, update a CRM record, and send a Slack notification, all as part of a single workflow.

Hallucination

When an AI system generates information that sounds plausible but is factually incorrect or entirely fabricated. Understanding hallucination is essential for professionals using AI, as it underlines the importance of verification.

Example: An AI might confidently cite a legal case that does not exist, or attribute a quote to the wrong person. This is why human review remains essential.

Context Window

The amount of text an AI model can process in a single conversation or session. A larger context window allows the AI to work with longer documents and maintain coherence across extended interactions.

Example: Claude has one of the largest context windows available, allowing it to read and analyse entire reports, contracts, or research papers in a single session.

Tools

Claude

An AI assistant and agent created by Anthropic, widely regarded as the most capable AI for complex reasoning, long-form writing, and agentic tasks. Claude is designed with a strong focus on safety and reliability, making it particularly well-suited for professional use cases.

Example: A consultant uses Claude to research a client's industry, synthesise findings from multiple sources, and produce a structured briefing document, a task that previously took a junior analyst half a day.

Anthropic

An AI safety company and the creator of Claude. Founded in 2021 by former members of OpenAI, Anthropic focuses on building AI systems that are safe, reliable, and interpretable. It is one of the leading AI research organisations in the world.

Example: Anthropic's research into AI safety and interpretability informs how Claude is designed to handle complex, sensitive professional tasks.

Practical Skills

Prompt

The instruction or question you give to an AI system. A prompt can be as simple as a single question or as detailed as a multi-paragraph brief specifying the role, task, context, and desired output format. The quality of your prompt directly determines the quality of the AI's output.

Example: "You are an experienced HR consultant. Review the following job description and suggest three improvements to attract more senior candidates. Format your response as a numbered list with a brief explanation for each suggestion."

Prompt Engineering

The practice of writing clear, structured instructions for AI systems to produce useful, accurate, and well-formatted outputs. Despite the technical-sounding name, prompt engineering is fundamentally a communication skill, the ability to describe what you want clearly and specifically.

Example: A marketer who consistently gets high-quality first drafts from Claude has developed strong prompt engineering skills, not by learning to code, but by learning to brief AI the way they would brief a skilled colleague.

System Prompt

A set of instructions given to an AI agent at the start of a session that defines its role, behaviour, constraints, and goals. The system prompt is the foundation of any AI agent, it tells the agent who it is, what it is trying to achieve, and how it should behave.

Example: A system prompt for a legal research agent might specify: "You are a legal research assistant specialising in UK contract law. Your role is to identify relevant case law and summarise key principles. Always cite your sources and flag any areas of uncertainty."

No-Code AI

Building AI agents and workflows using visual interfaces and plain language instructions rather than writing programming code. No-code tools make agentic AI accessible to non-technical professionals.

Example: Using a platform like n8n or Make, a marketing manager can build an AI agent that monitors social media mentions, analyses sentiment, and creates a weekly report, without writing a single line of code.