A current-state assessment of your systems landscape: where data is transferred manually, where media breaks occur and where paper is used in parallel.
process digitalisation
Process digitalisation and AI: first the process, then the tool
Process digitalisation before automation
My position is simple: first the process, then the tool. A process with media breaks, duplicate data entry and unclear responsibilities does not become better through AI, only faster at being poor. Process digitalisation therefore means understanding, streamlining and standardising first, then automating.
AI is here to stay, and the possibilities are considerable. My focus lies one step earlier, with the processes. For implementation, I am happy to work with my network, and if we decide together that AI is the next step, I will take that step with you. Talk to me.
Sound familiar?
The systems are there, but the processes between them are not.
What you get
First the processes, then the tools.
Transparency about AI use within the company: who uses which tool, for what purpose and with which data.
Streamlined, standardised and documented processes as the basis for any automation.
An organised overview of potential applications, prioritised with your specialists by benefit and effort.
An AI workshop for managers: assess potential and risks before decisions are made.
A decision basis for leadership: what is worthwhile, what can wait and where specialists are needed.
Access to specialists from my network when technical implementation is due.
Support for implementation on the process and change side.
Frequently asked questions
Process digitalisation and AI: common questions
Where does AI help in production and administration?
In administration, wherever a large volume of similar work arises: reading and assigning documents, answering recurring enquiries, searching documents, and preparing texts and summaries. In production, it can help analyse machine and quality data, recognise images in inspection and predict failures. The rule is the same in both areas: the more frequently a process occurs and the more uniform it is, the more likely it is to be worthwhile. For individual cases with many exceptions, the effort is usually greater than the benefit.
How can you tell whether a company is ready for AI?
By three things: the process is documented and runs the same way everywhere; the relevant data is complete and can be found in one place; and someone makes the decision and remains responsible afterwards. If any one of these is missing, implementation becomes expensive.
What does digitalisation achieve if AI is not yet relevant?
A great deal, with immediate effect. A process that no longer requires duplicate entry saves time every day, regardless of whether AI is added later. Forms, interfaces and workflows may be unspectacular, but they solve much of what prompts calls for AI. They also create exactly what AI will need later: clean data and a process that runs consistently everywhere.
Where do you start if AI is already being used in the company?
With an inventory. First put the facts on the table: who uses which tool, for what purpose and with which data? This reveals both the useful applications worth expanding and the risky ones that need rules. A ban merely shifts use onto private devices, where nobody can see it.
Why do companies need AI governance?
Because otherwise everyone decides individually what to enter into a tool. Governance here need not be elaborate: an overview of permitted tools, which data may and may not be entered, who approves use and who is consulted in case of doubt. It also defines that results must be checked before leaving the company. This is developed in a facilitated workshop and approved by the company together with its own IT function.
Why do AI introductions fail?
Usually not because of the technology. The most common reason is a process that runs differently in every department, forcing the solution to be adapted for every variant and making it expensive. The second is the data: no tool can analyse what has not been recorded cleanly. The third is a pilot with no route into daily work: it functions, but nobody is responsible for embedding it. That is why process work comes before the tool.