AI Agents

An autonomous software system that perceives its environment, reasons about goals, and takes actions without continuous human intervention.

Perception (sensing environment), Reasoning (planning and deciding), Action (executing tasks), and Learning (improving over time).

An AI agent can autonomously plan, use tools, maintain state across interactions, and take multi-step actions, whereas a chatbot typically only responds to individual prompts.

A cycle where the agent observes the environment, thinks/reasons about what to do, acts on that decision, then observes the result and repeats.

The ability of an agent to invoke external tools (APIs, code interpreters, databases, web browsers) to accomplish tasks beyond text generation.

A multi-step process where an AI agent autonomously plans, executes, evaluates, and iterates on tasks to achieve a goal.

A framework combining Reasoning and Acting: the agent thinks step-by-step (chain of thought) and interleaves actions (tool calls) with observations.

Connecting the agent's responses to real-world data sources, tools, or verified information to reduce hallucination.

The degree to which an agent can make decisions and take actions without requiring human approval at each step.

A system design where one AI agent handles all tasks, reasoning, and tool use within a workflow.

A system where multiple specialized AI agents collaborate, each handling different aspects of a complex task.

The coordination layer that manages how multiple agents communicate, delegate tasks, share context, and combine their outputs.