AI Agents
AI agents are LLM-powered systems that can plan complex tasks autonomously, use external tools, and learn from previous steps. Unlike simple language model queries, agents execute multi-step workflows and interact actively with their environment, from databases and APIs to other agents.
Contents
What is an AI agent? Distinction from classic LLMs and typical use cases.
The three core building blocks: planning, tool use, and memory.
How agents invoke external tools. The agent loop in detail.
Orchestrator-subagent architectures for complex, long-horizon tasks.
How instructions, tool definitions, and memory produce reliable agents.
Last updated on