Two founding pilots.

An AI system can be authorised to act. A tool can execute correctly. None of these establishes where a consequential figure came from.

For the figures in scope, each pilot establishes whether the source values and the computation behind each figure can be evidenced; whether the result can be independently reproduced; and whether the system declines to release a figure when the evidence required to support it is unavailable.

One workflow. One class of consequential figure. Success criteria agreed before we begin.

This is not a broad AI transformation project. It is a deliberately bounded technical pilot.

Certanum funds the initial pilot work.

In return we look for a serious technical partner, a named internal sponsor and, where possible, permission to publish an anonymised methodology and findings.

Where the pilot meets its agreed criteria, we agree a route to a commercial phase. The scope, success criteria, data handling and what each party contributes are set out in writing before any work begins.

Four questions.

Selection reflects fit with the scope above, not the order in which expressions of interest arrive. Every expression of interest is read by the founders.

Opens your email client with the four questions filled in, addressed to pilots@certanum.com. Please do not include confidential or production data in an expression of interest.