L

Learn AI

Track progress · learn offline

Open

Lesson 6 of 7

From workflow to agent

7 min read

Every flow so far ran the exact steps you drew, in the exact order. But what if you don't know the steps ahead of time — because the right next move depends on what just happened?

When the path isn't fixed

A workflow is a fixed path: you decide every step in advance and it runs the same way each time. Perfect when the job is predictable. But some jobs aren't — the right next step depends on what you just found out. That's where an agent comes in: you hand it a goal, not a script.

An agent gets a goal, not a fixed list of steps. It loops — think, act, observe — and picks its next move from what it just saw.

Fixed flow vs. deciding on the fly

Watch the difference. A rigid workflow told to "book 7pm" would just fail when 7pm is full. The agent notices 7pm is gone and tries 8pm on its own — nobody wired that branch in advance. It's the gap between following a recipe and having a cook who adapts.

Workflows are reliable and repeatable; agents are flexible but less predictable. Use a workflow when you know the steps, an agent when you don't.

Agents are powerful, but they can wander or make mistakes. Keep a human in the loop for anything costly, risky, or hard to undo.

The shape of it

When is an agent a better fit than a fixed workflow?

Continue in the app

Take the whole Automation 101 course — tracked

Get your personalized path, progress and streaks in the app — this lesson and every next one, in order.