English companion article · Published

Understand the moving parts

The episode begins with the difference between chatbots and agents, then looks at how an agent works. The key idea is that a model can sit inside a larger process: it receives a goal, uses available information and tools, and contributes to completing a task. The surrounding system matters as much as the conversational interface.

From one agent to a workflow

The guide introduces multi-agent workflows before discussing ways to build an agent. It includes no-code approaches and an explanation of n8n. These sections show that building an agent is partly about connecting useful capabilities, not simply finding a clever prompt.

More components also mean more behaviour to understand. Start with a small workflow whose inputs and outputs you can inspect before introducing several agents or many external connections.

Find a practical use case

The closing section looks at business opportunities. Begin by identifying recurring work that has clear information sources and a recognisable result. Then ask whether an agent is actually useful for that task. A simple fixed workflow may be sufficient in some cases; other tasks benefit from more flexible decisions. The video offers a conceptual introduction before the longer Agentic AI course.

← Explore all articles & videos