Insights
Findings from building, not AI explainers.
What we write here comes from working with these systems - with figures, tables and the places where it tips over. General explainers about artificial intelligence already exist in quantity.
The bottleneck is never the building
Why an audit comes first here, and not a quotation.
ReadPerspectiveWhat AI can actually do in a business today
Three stages of build-out, and why businesses that grew organically stand to gain most.
ReadInstrumentHow to tell whether a process suits AI
Five questions you can answer without us - and one you cannot.
ReadPerspectiveWhat an AI project costs in a mid-sized business
What the effort depends on, and from when a project pays for itself.
ReadPerspectiveWhy an assistant does not yet automate the process
The difference between a tool someone operates and a routine that runs by itself.
ReadInstrumentWhat an agent is allowed to decide
Four levels of independence and the rule by which we assign them.
ReadPerspectiveWhat AI consulting delivers - and where it stops
Why a recommendation without implementation usually changes nothing in a mid-sized business.
ReadOn the selection
Why there are no success stories here
Every piece here may only claim what we know from the work: from building, from running our own systems, from using these models daily. Client results are not in there.
Invented case figures would be the easiest thing to put here and the first thing an attentive reader recognises as a claim. We would rather write about the questions we can answer - and add the figures once they come from real projects.
Let us start with a conversation.
You tell us where your time goes. We put it in writing: what is worth automating, and roughly what that costs.
