7 terms
Agent memory is the set of mechanisms that let an AI agent keep, retrieve and update information across steps, sessions and tasks beyond its context window.
Agent orchestration is the control layer that sequences, routes and supervises the steps, tools and sub-agents of AI agents in a multi-step task.
Agentic AI describes AI systems that pursue goals with autonomy, planning steps, calling tools and acting on results with limited human direction.
An AI agent is a system that pursues a goal by observing its environment, planning steps and acting through tools in a loop, with limited human…
MCP (Model Context Protocol) is an open standard that lets AI agents and assistants connect to external tools, data sources and prompts in a uniform way.
A multi-agent system is an AI architecture in which several agents with separate roles, tools and context cooperate or compete to complete a task.
Tool calling lets a language model request external functions with structured arguments, so AI agents can fetch data and take actions beyond text generation.