What is an AI agent?

What it is

An AI agent is a system where a language model decides its own sequence of actions — which tools to call, what to check, when it is done — rather than following a hand-coded script. This is different from a fixed "workflow" where a developer predetermines every step in advance.

Why it matters for you

Agents let you automate work whose steps you can't fully predict in advance — research, multi-system tasks, customer support triage — because the system adapts its plan as it goes. That flexibility is also the cost: less predictable, harder to test, and pricier per task than a simple scripted workflow, so it's worth reaching for only when the task genuinely needs judgment calls, not for anything a fixed pipeline could handle.

Common failure mode

Teams reach for an "agent" when a simple deterministic workflow (call step 1, then step 2, then step 3) would be cheaper, faster, and easier to debug — autonomy is added where it isn't needed, and the result is a slower, less predictable system for no real benefit.

This is exactly what we work through in a session.