AI Agents
AI agents are software systems that use AI models to interpret information, decide on actions, and carry out tasks toward a goal, often by calling tools, data sources, or other applications.
An AI agent is the working unit of agentic AI. Where agentic AI describes the broader capability of acting toward goals, an agent is a specific system built to do a job, such as resolving support tickets or reconciling invoices. Most agents combine a language model with instructions, access to tools, and some form of memory. Some work alone, while others work as part of a team of agents.
Agents are often confused with chatbots and robotic process automation (RPA). A chatbot mainly answers questions in a conversation. RPA follows fixed, scripted steps. An agent can decide which steps to take based on the situation, which makes it more flexible but also means it needs clear limits on what it can access and do. Well-designed agents have a narrow purpose, limited permissions, and a person responsible for their results. They also keep a record of what they did, so their actions can be reviewed.
Examples
A support agent that reads a ticket, checks account data, and drafts a resolution.
A coding agent that writes and tests changes to a codebase for review.
A research agent that gathers sources and summarizes findings on a topic.
An operations agent that matches invoices to purchase orders and flags exceptions.
