AI Agents and Multi-Agent Systems · Lesson 1
Foundation of Agents
An AI agent is a system where an LLM decides what to do next in a loop: observe, think, act with a tool, observe the result, and repeat until the goal is reached.
Agents differ from fixed workflows. A workflow follows steps you coded; an agent chooses its own steps at runtime. Use workflows when the path is known and agents when tasks are open-ended.
Autonomy brings risk: loops, cost blow-ups and wrong actions. Set step limits, budgets and human approval for important actions.
Key points
- Agent = LLM in a loop choosing actions
- Workflows are predictable, agents are flexible
- Add step limits and budgets
- Keep humans in the loop for risky actions