Agents Get the Attention. Workflows Get the Work Done.

Somewhere between prompting and agents, the industry skipped a step. We went from “write me a good prompt” to “build me an agent” without much time in between. Somewhere in that transition, we stopped talking about AI workflows.

An agent gives the model tools and room to figure things out. It can handle situations you did not anticipate and combine tools in ways you did not design. It sounds good in a pitch. Getting it to work reliably is harder than it looks.

A workflow is a predefined series of steps. Each one is focused and testable. The model handles one subtask at a time, which keeps output precise. When something breaks, you know where to look. Agents are the right call when the problem is genuinely open-ended. Most problems are not.

Start with workflows. Reach for agents only when the problem genuinely requires it. Predictability is the product — it is what users are paying for.

Originally published on Substack.