One workflow. Thirty days. A fixed price, and you keep what we built.

• The workflow, written down and running on your systems.
• The rules, versioned and approved by your team.
• The records: every check and decision from the Proof, with the rule and version it used.
• The test set: past records your team signed off, which every candidate was measured on.
• The comparison: how each candidate did against the bar from day 5.
• The evidence pack: the records and test results, traced to your own rules, plus a pack showing how the workflow meets the rules in one country or region. Mapping to industry or safety standards isn't included unless we scope it.
• A three-year cost and exit map: what it costs to run, and how to leave every dependency.
• A short briefing for your board or leadership.

You keep the data, the rules, the tests and the documents; nothing is locked to us. The models, test sets, instructions and documents are yours when we leave. In a Proof everything runs on your systems and we never hold your data. The one exception is a comparison against an outside hosted model, and only where your data policy allows it.

Can this go through our procurement?

A Proof is a fixed-price statement of work with written scope, deliverables and acceptance: the bar. Tell us on the first call what your procurement needs, such as vendor registration, a security questionnaire or a data-processing agreement, and we'll say plainly what we can provide.

What do we have if we stop working together?

It all runs on your systems, and you keep what we built: the data, the test set, the instructions, any models and the documents, including the written workflow and rules.

"Just automate this. We don't want AI."

We'll take that brief. If a simple rule does the job, we'll use a rule and tell you so. Where rules fall short, such as reading free-text notes, a model does that part, tested first on your own past records, and your team reviews what it flags.

"Why touch something that works?"

You don't have to. We start beside it, and the old way keeps running until your team chooses otherwise.

"Will AI kill my business tomorrow?"

Not if you start small and learn on your own work: one workflow, on rules your team approved. You find out what it can and can't do for you from your own records, not from headlines.

"Will my people push back?"

They stay the ones who decide. They get only the cases that need them, and they switch when they trust it.

"Our inspections are on paper. Where do you start?"

With what's already captured: registers, spreadsheets, scanned sheets. In a Proof we read those, and never change them. If the records exist only on paper, the first job is getting them into a form software can read, and we'll tell you on the first call whether that fits in 30 days.

"Can't our team just use a chat assistant?"

It helps one person write faster. It doesn't check work against your rules or record who decided what, so someone still has to check everything it produces.