AI glossary
AI Glossary for Business Leaders
Plain definitions of the AI terms that come up in strategy meetings and vendor pitches, from a team that builds these systems.
Why Operator Intelligence keeps an AI glossary
AI vocabulary moves faster than most teams can track, and vendors use the same words to mean different things. One vendor's "agent" is another's chatbot; "orchestration" can mean a workflow tool or a lead model directing others. When a leadership team and a vendor use a word differently, projects get scoped wrong.
Operator Intelligence is an AI enablement consultancy for lean companies, and we use these terms every day in scoping calls, builds and team training. This glossary is how we define them: in plain language, with the distinctions that matter for a buying or build decision, and with client examples where we have them.
Where a fact depends on a date, such as who governs a protocol, the entry says so.
AI terms A–Z
Sixteen terms, each linked to its full entry.
Agentic AI
AI that pursues a goal by planning steps, using tools and adjusting to results, within limits people set.
Learn moreAgentic automation
Process automation where AI agents handle the judgment steps and fixed workflows run the rest.
Learn moreAI agent orchestration
The layer that routes work between AI agents and tools, passes context and enforces approvals.
Learn moreAI center of excellence
A small cross-functional group that sets AI standards, prioritizes use cases and shares what works.
Learn moreAI enablement
The work of getting a company's people, processes and tools to use AI in daily operations, with measurable results.
Learn moreAI literacy
The baseline knowledge people need to use AI sensibly, check its output and protect data.
Learn moreAI maturity model
A staged framework for judging how far an organization has come with AI, from early experiments to AI built into operations.
Learn moreAI ROI
The return on an AI investment: hours saved, errors avoided or revenue gained, set against build and running costs.
Learn moreClaude Skills
Folders of instructions, scripts and resources that Claude loads when a task calls for them.
Learn moreGenerative engine optimization (GEO)
Making content more likely to be cited and described accurately by AI answer engines such as ChatGPT.
Learn moreHuman-in-the-loop
A design in which a person reviews, approves or corrects an AI system's output before it takes effect.
Learn moreModel Context Protocol (MCP)
An open standard, introduced by Anthropic in 2024, for connecting AI applications to external tools and data.
Learn moreMulti-agent systems
Groups of specialized AI agents, each with its own role and tools, coordinated on tasks too broad for one agent.
Learn morePrompt library
A shared, maintained collection of tested prompts and instructions a team reuses for common tasks.
Learn moreRetrieval-augmented generation (RAG)
A technique that retrieves relevant passages from your documents and has a language model answer from them.
Learn moreShadow AI
The use of AI tools, accounts or features at work without the company's approval or oversight.
Learn moreAbout the AI glossary
Who writes the Operator Intelligence AI glossary?
The Operator Intelligence team, led by co-founders Chad Westleigh and Luke Freeman. Technical facts are checked against primary sources and dated where they may change.
What is the difference between agentic AI and generative AI?
Generative AI creates content such as text, images or code in response to a prompt. Agentic AI pursues a goal by planning steps and taking actions with tools, usually with a generative model inside. See the agentic AI vs generative AI comparison.
Where should a business leader start?
With the terms behind your next decision: AI enablement and AI literacy for a company-wide plan, MCP and human-in-the-loop for connecting AI to your systems, shadow AI if staff already use their own tools.
Can you help us apply these ideas?
Yes. Operator Intelligence helps lean companies put them to work through AI enablement, custom AI solutions and team training. Our case studies show agentic workflows, MCP and multi-agent design in client builds.
Turn AI vocabulary into a working plan
Knowing the terms is step one. Find out where your company stands with the free, five-minute AI readiness assessment, or talk to us about the first workflow worth changing.
Book a scoping callSources
- Anthropic, "Introducing the Model Context Protocol" (2024): Anthropic introduced the Model Context Protocol on 25 November 2024 as an open standard.
- Anthropic, "Introducing Agent Skills" (2025): Agent Skills are folders of instructions, scripts and resources that Claude loads when needed (launched 16 October 2025).