Guides · Choosing an approach

Do you need an AI agent, or just automation?

Short answer

If the rules are stable and the inputs are structured, plain automation is faster to build, cheaper to run and easier to trust — and you should use it. A language model earns its place only where the input is unstructured language and no one can write the rule. Most working systems are automation with a model doing one specific job inside them.

Rules vs judgement
The single question that decides which one you need
Cost per run
A deterministic step costs fractions of a cent; a model call does not
Usually both
Automation as the frame, a model inside one or two steps
By Max Pochinsky · Updated August 2026 · Written for people scoping a project, not for search engines

What each one actually is

Automation means steps you specified: when an invoice arrives, pull these fields, check them against the purchase order, post to the ledger, notify the buyer if the totals differ. It runs the same way every time. When it breaks, it breaks loudly and in a place you can point at.

An agent means a language model deciding what to do next within limits you set. It reads something unstructured — an email, a call transcript, a contract — decides which of your tools to use, and produces an outcome. It handles input nobody could enumerate in advance, and it does not run the same way every time.

The question that decides it

Ask the person who does the work today to write down the rule they follow. Not the summary — the actual rule, including the exceptions. What happens next tells you almost everything.

01

They can write it down: use automation. It will be cheaper to build, cheaper to run, and you will be able to prove what it did.

02

They cannot, because it depends on what the message says: that is the model's job — reading language and turning it into something structured your automation can act on.

03

They cannot, because nobody has decided: neither approach helps. The decision has to be made by a person before any system can execute it, and pretending otherwise is how pilots die.

Why the answer is usually both

In practice the argument is misframed. A support system is not an agent; it is a workflow with a classification step where a model reads the ticket, and a retrieval step where a model drafts an answer from your documentation. Everything around those two steps — routing, permissions, escalation, logging, the CRM write-back — is ordinary automation, because it should be.

That shape is deliberate. The deterministic frame is what makes the system auditable and cheap; the model is confined to the parts that genuinely need judgement about language. When a project goes wrong, it is very often because a model was handed a job a rule could have done.

What changes when a model is in the loop

01

Cost per run stops being negligible. Deterministic steps cost almost nothing. Model calls have a real unit price, so volume that was irrelevant to your infrastructure bill becomes a line item worth forecasting.

02

The same input can produce different output. That is the point, and it is also the problem. It means you need evaluation — a set of real cases with known good answers — rather than a test that passes once.

03

Failure gets quieter. Automation fails with an error. A model fails with a fluent, wrong answer, which is why citations, confidence thresholds and a human path for low-confidence cases are not optional extras.

04

Latency goes up. Seconds instead of milliseconds. Irrelevant in a support reply, occasionally decisive in a checkout flow or an API someone else waits on.

How to decide this week

Take one process. Split it into steps and mark each step as rule or judgement. If every step is a rule, you do not need a model and any vendor telling you otherwise is selling. If one or two steps are judgement about language, that is where a model goes — and the rest stays deterministic.

If most steps are judgement and they are not about language but about your business — pricing, credit, hiring, who to fire from a supplier list — that is not an automation project yet. It is a decision-making project, and it needs a person to set the policy first.

Follow-up questions

What people ask next.

No — the difference is who chooses the next step. In automation you chose it when you built the system. With an agent, the model chooses at run time within limits you set. That flexibility is exactly what you want for messy input and exactly what you do not want for a payment run.

Still unsure whether your process is worth automating?

Bring us the process. We take it apart with you at no charge and give you a straight answer, including when the answer is no.