AI process automation for manufacturers, starting with the process

Automating a poor process gives you a faster poor process. So we begin where Lean begins: map the process, remove what adds nothing, and agree a standard. Only then do we look at what a tool could take over.

A newer part of our work

This is a newer part of our work, built on 25 years of improving the processes that automation depends on.

We start small: one process, mapped and standardised. Then we automate only the steps where it clearly helps, the repetitive, rule-based work, and leave judgement and decisions with your people.

What you can ask us for

  • Process mapping. A clear picture of how the process runs today, built with the people who run it.
  • Automation potential. Which steps are worth automating, which are not, and why.
  • Basic automation, implemented. Simple workflows built with tools such as n8n and ChatGPT.
  • AI risk assessment. A structured, non-legal assessment of your AI use against the EU AI Act. It does not replace legal advice.
  • AI governance. Clear rules, roles and approvals for how AI is used in your organisation.

Standardise first, then automate: two examples

These come from our Lean projects. They are process automation, not AI projects, and they show the order we work in.

  • Back office: once the process for renewing temporary contracts had been mapped and standardised, it moved from manual email chains and file transfers to a largely automated workflow.
  • Equipment manufacturer: once rework had standard codes, a daily dashboard replaced a manual monthly data pull.

Further examples are in development and will be added here.

Have a process in mind?