Our approach.

The workflow described in a procedure is almost never exactly the one that exists in practice.

We therefore start with the people who do the work.

We seek to understand how a task unfolds: what triggers it, the information used, the decisions made, the checks carried out, the exceptions encountered and the judgement calls that are written down nowhere.

A single operation may look simple on paper, yet in reality involve several people, several files, several systems and a considerable amount of human judgement.

It is often in this gap, between the workflow as described and the work as it is actually done, that the opportunity lies.

We then turn to what the systems have recorded.

What data exists? Where is it stored? Which systems are involved? How does information flow? Where are the breaks, the duplicates, the double data entry and the approvals?

The systems can tell a different story from the one people tell.

Information may sit in an ERP yet be systematically checked in a spreadsheet. Data may be officially available yet remain unusable because it has to be reprocessed manually. A workflow may appear automated when, in fact, a person intervenes at several stages to keep it running.

Understanding this reality allows us to build around what already exists, rather than asking the company to replace it.

Communication tools are another source of inputs: emails, WhatsApp, Slack, messages, notifications, attachments and so on. Through them, information, requests, decisions, instructions and sometimes signals enter the work without being structured in the company’s systems.

Combining these three dimensions makes it possible to understand what must change and where intelligence must intervene.

Our work is then to build the system suited to this reality, without asking the company to become something it is not.