Glossary

The AI agent terms that show up everywhere, explained in two sentences.

A program that uses a language model in a loop to decide actions, call tools, and work toward a goal with little step-by-step instruction.

Multi-Agent System

A setup where several specialized agents work together, each handling part of a task, instead of one agent doing everything.

The mechanism that lets a model invoke external functions, like a search or a calculator, instead of only generating text.

Retrieval-Augmented Generation (RAG)

A technique that fetches relevant documents and feeds them to the model so its answer is grounded in your data, not just its training.

Human-in-the-Loop

A pattern where an agent pauses for a person to approve, edit, or reject a step before continuing, used for consequential actions.

State Persistence

The ability to save where an agent run is and resume it later, surviving crashes and spanning separate requests.

The control logic that decides what an agent does next: which tool to call, which agent runs, when to branch, and when to stop.

Agent Observability

Seeing every model call, tool call, and decision an agent made, so you can debug failures and improve behavior.