You already have AI. You do not have the results yet.

A Ukrainian company of 150 people delivering digitalisation services across the country bought licences for a project management system. The system was configured properly, but almost nobody used it; status was still assembled from mail threads the evening before the management report, and everyone noted who promised what to whom their own way. When we asked why people were not using the system, we heard that the tool was inconvenient and brought no value. It is the most common picture we see: the tool is there, and it has nothing to work with.

We did not recommend a different tool. We ran a maturity assessment, proposed a data-driven approach to running the portfolio and a few practices, built the context the system was missing, and worked with the PMO until the system became the place where the work lives rather than the place data gets entered just before reporting. In six months, from the end of the first quarter to the start of the fourth, maturity in the project-management-with-automation domain went from 1.88 to almost 3 (2.85) on a 0 to 5 scale, right up against the level of systematic use. Then came an audit by one of the Big Four, on which the company's further funding depended. The PMO pulled the information and the answers, the so-called documented evidence of practice, out of the system in a structure the auditors accepted. Audit passed, funding secured.

We did not recommend a different tool. We ran a maturity assessment, proposed a data-driven approach to running the portfolio and a few practices, built the context the system was missing, and worked with the PMO until the system became the place where the work lives rather than the place data gets entered after the fact, just before reporting. In six months, maturity in project-management grew from 1.88 to almost 3 (2.85) on a 0 to 5 scale, right up against the level of systematic use. Then came an audit by one of the Big Four, on which the company's further funding depended. The PMO pulled the information and the answers, the so-called documented evidence of practice, out of the system in a structure the auditors accepted. Audit passed, funding secured.

We tell this story not because it is pretty, but because it has a number before and a number after. We are ready to measure improvements like these for you too.

Your AI runs on the same models as your competitor's, and the difference is not the model but what it knows about your company. Who promised what to whom, why exactly that decision was taken, which deadline is contractual and which is just a hope. Without that, AI writes plausible text with wrong specifics, managers check it, then stop using it, and adoption quietly dies. A system without context is like a kitchen without groceries: the stove is there, dinner is not.

The layer where AI spend turns into results is the management layer, and its raw material is context. That is what we build, in your tools, with your managers, and that is why our work is measured, not “delivered”.

The first path: buy a tool and hire an integrator. You get a configured system and empty dashboards, because the change-management method and the building of historical context were most likely never considered. The second path: give a smart PMO lead ChatGPT and hope. For three months it looks good, then it turns out the answers are plausible, the details are wrong, and you have no evidence to show a board or an auditor. The third path: do nothing and keep relying on the same strong people. It is the most expensive one, the bill just arrives later, as a quarter of recovery after every departure.

We offer a fourth: measure first, then build on what you already have, and show the numbers before and after. The first step costs two weeks of your time, and if we are not the right fit we will say so ourselves, because turning down work we do not believe in is the cheapest thing we have ever done for a client.