Is this a trial, or do we keep what you build?

You keep it. The pilot is the first production build, not a demo that gets thrown away. It runs on your systems with your data, and if you stop after 30 days it keeps working. What is limited is the scope, not the result.

Why $30,000 for 30 days?

Because the 30 days include the parts that usually get skipped: mapping the process, recording a baseline you can measure against, defining escalation rules in writing, and testing against real cases. Those are the parts that decide whether an agent survives contact with your operation.

What if it does not work?

Then we say so at the closing measurement, with the numbers next to the baseline. We would rather lose the follow-on work than have you run something that does not pay for itself. That is also why the scope is one process — a failure costs you 30 days, not a year.

Do we have to change our software?

No. The agent connects to what your team already operates: phone system, CRM, calendar, email. The pilot is deliberately designed so that no migration is needed, because a migration would consume the 30 days on its own.

One of the three founders, directly. There is no account manager between you and the person building the agent, and no handover to a delivery team you have not met.

The offer

The 30-day fixed-price AI agent pilot

One process, 30 days, $30,000 agreed before we start. You get a working agent inside the systems you already run, escalation rules in writing, and a closing measurement against a baseline recorded on day one. If the numbers do not justify keeping it, we say so.

By Gorden WübbeAutomation, agents and search visibilityUpdated 10 August 2026

Why a pilot and not a project

Most AI agent work fails the same way. The scope grows, the timeline slips, and eighteen months later there is a platform nobody uses and an invoice nobody can justify. It fails because the first step was never small enough to finish.

Thirty days on one process is small enough to finish. That is the entire design principle. The constraint is the product.

The pilot is a build, not a demo

This is worth being precise about, because “pilot” often means a throwaway prototype. It does not here. The agent runs on your systems with your data, and if you walk away on day 31 it keeps working. What is limited is the scope, not the quality — and not your ownership of the result.

What the 30 days actually contain

Days 1–5: the process, measured

We pick one process that is costing you time and write down how it works today — including the parts that only exist in someone’s head. Then we record a baseline: how many, how long, how often it goes wrong. Without that number, nothing at the end of the month can be proven.

Days 6–20: the build

The agent is built inside your existing stack. Phone system, CRM, calendar, email — whatever the process already touches. In parallel we define, in writing, the rules for handing a case to a person: which situations, to whom, with what context.

Days 21–27: real cases

The agent runs against real traffic with a person watching. This is where the awkward cases surface — the caller who asks two things at once, the request that does not fit any category. Each one either gets handled or becomes an escalation rule.

Days 28–30: the closing measurement

The same measurements as day one, side by side with the baseline. Plus a recommendation that we commit to in writing: keep it, extend it, or switch it off.

What is in scope and what is not

Included	Separate quote
One process, end to end	A second or third process
Connection to the systems that process already uses	Replacing or migrating a system
Escalation rules, defined in writing	Rebuilding the process itself before automating it
Baseline and closing measurement	Ongoing operation beyond day 30

The right-hand column is not upselling. It is the list of things that would quietly consume the 30 days and leave you with nothing finished.

Why the price is fixed

$30,000, agreed before anything starts. Not “starting from”, not time and materials. Every agency in this market quotes a range, and every range is read by the buyer as “it will be more” — usually correctly.

A fixed number is only honest if the scope behind it is fixed too, which is why the scope is one process and why the table above exists. Both halves have to hold, or neither means anything.

What happens on day 31

Three options, and we tell you which one we would choose in your position: keep it running as it is, extend it to a second process under a separate quote, or switch it off. Nothing renews on its own, and there is no contract that makes leaving awkward.

If the closing measurement does not justify keeping it, that is the recommendation you get. We would rather lose the follow-on work than have you running something that costs more attention than it saves.

What people ask before they commit

Is this a trial, or do we keep what you build?: You keep it. The pilot is the first production build, not a demo that gets thrown away. It runs on your systems with your data, and if you stop after 30 days it keeps working. What is limited is the scope, not the result.
Why $30,000 for 30 days?: Because the 30 days include the parts that usually get skipped: mapping the process, recording a baseline you can measure against, defining escalation rules in writing, and testing against real cases. Those are the parts that decide whether an agent survives contact with your operation.
What if it does not work?: Then we say so at the closing measurement, with the numbers next to the baseline. We would rather lose the follow-on work than have you run something that does not pay for itself. That is also why the scope is one process — a failure costs you 30 days, not a year.
Do we have to change our software?: No. The agent connects to what your team already operates: phone system, CRM, calendar, email. The pilot is deliberately designed so that no migration is needed, because a migration would consume the 30 days on its own.
Who do we work with?: One of the three founders, directly. There is no account manager between you and the person building the agent, and no handover to a delivery team you have not met.
What happens on day 31?: Three options, and we tell you which one we would pick. Keep it running as it is, extend it to a second process under a separate quote, or switch it off. Nothing renews automatically.

Start with one process.

Tell us which one costs you the most time. Thirty minutes is usually enough to say whether it fits.

Book a scoping call