AI Agent Architecture
An AI agent is an LLM that can reason, plan, use tools, and take actions to complete a goal.
Core Components
| Component | Role |
|---|
| LLM (brain) | Reasoning and decision making |
| tools | Functions the agent can call (search, code execution, APIs) |
| memory | Short-term (context window) + long-term (vector store) |
| planner | Breaks down complex goals into steps |
| executor | Runs the planned steps and handles tool outputs |
Agent Patterns
- ReAct: Reason + Act — the model alternates between thinking and taking actions
- Plan-and-Execute: First create a full plan, then execute each step
- Multi-agent: Multiple specialized agents coordinated by an orchestrator
Useful Phrases
- "The agent uses a ReAct loop — it reasons about the next action, executes it, then observes the result."
- "We have a orchestrator agent that delegates subtasks to specialized worker agents."
- "Long-term memory is stored in a vector database and retrieved when relevant."