Who it is for
- A team that does checks or verifications by hand every day: orders, invoices, reconciliations, quality, request triage.
- A task that comes back at least a few hundred times a month.
- One person in the team, ideally already using an AI assistant, ready to describe their task.
How it works
| Week | What happens |
|---|---|
| 1. Scope | We pick the task with you. The person who knows it describes it to their assistant. We measure the starting point: time spent, errors, rework. Your IT approves access. |
| 2 and 3. Compile | Maurice builds the tool. We try it on real cases; the person adjusts their own rules, in plain language. |
| 4. Measure | The tool runs for real. We measure cost per run, time saved, correct results. You decide what comes next. |
| 5 and 6, optional | Consolidate, or try a second task. |
What you get
- A tool that runs, on your task.
- The before-and-after measurement, on your own data.
- Permissions set up: who uses the tool, who approves it.
- A written review, with its limits.
What is agreed before we start
- Scope, price and success criteria.
- What goes through an AI model, and where.
- Exit terms: you keep your tools and your data.
What it costs
The pilot is on quote, depending on the scope of the task. After that, Maurice costs the price of its plan: see pricing.
Frequently asked questions
Do we need a tech team? No. You need one person who knows the task, and one IT contact to approve access.
Where will the tool run? On the person's computer, on your server, or on a machine we host. We decide during scoping.
What if the results are not good? You will know in four weeks, with numbers, and you stop. That is what the pilot is for.