Glossary.

Short definitions of the terms we use when we talk about bringing AI into how a company works, and where each one sits in the four stages.

Acceptance criteria
What acceptance criteria are, with examples and common formats, and why AI work needs them: they make done mean the same thing for everyone.
Agentic workflow
Agentic workflow explained: AI agents that plan steps, use tools and act toward a goal. Examples, and how it differs from a multi-agent system.
AI knowledge base
What an AI knowledge base is, how it differs from a wiki, examples of what goes in it, and why AI needs one to know your business.
Context engineering
Context engineering in AI and LLMs: giving a model the knowledge it needs before it answers. How it works, an example, and how it differs from prompts.
Harness engineering
Harness engineering in AI: the frame a model works inside, with its task, tools and checks. What it is, whether it is real, and how it differs from context.
Human in the loop
Human in the loop meaning in AI: a person stays in the process where judgement is needed, while the routine runs on its own. Examples and why it matters.
LLM as a judge
LLM as a judge explained: using a language model to check AI output against criteria. How it works, where it helps and where it still needs a person.
Loop engineering
Loop engineering in AI: recurring work that runs in cycles, with a person asked only for decisions. What it is, how it works and what it needs first.
Model Context Protocol
Model Context Protocol (MCP) explained: an open standard that connects AI apps to tools and data. How it works, an example, and is MCP secure?
Prompt engineering
Prompt engineering explained: what it is, how it works, how it differs from context engineering, and whether it is a real job.
Retrieval augmented generation
Retrieval augmented generation (RAG) explained: an AI model looks up the right documents before it answers. How RAG works, an example, and RAG vs fine-tuning.
Second brain
What a second brain is, from personal notes to a company memory that AI can use. Second brain AI explained, and how it relates to a knowledge base.