Articles
Making intelligence operational.
Writing on connected manufacturing data, operational risk and the practical application of AI. Opinionated, specific, and written by people who have to ship this into a real operation.
Why manufacturing dashboards miss the decision
Reporting what happened is not enough when material, production and delivery dependencies are still moving. The gap is not visibility. It is the distance between a chart and a choice.
14 October 2026 · 9 min read
02 / Product thinkingWhat an order delivery control tower should actually do
The term is used loosely enough to mean almost nothing. Here is the difference between a late order report with a better name and a system that genuinely helps a team intervene.
7 October 2026 · 10 min read
03 / GovernanceBuilding trust in operational AI
Why source evidence, clear permissions and human approval matter from the first customer pilot, and what happens to adoption when they are missing.
23 September 2026 · 11 min read
04 / EconomicsThe exception rate decides whether operational AI pays
Accuracy is the number everyone asks about. The proportion of cases that complete without a human is the one that determines whether the system saves money or quietly costs more.
9 September 2026 · 8 min read
05 / ArchitectureOperational context: the thing your ERP does not have
Your systems hold the facts. What is missing is the layer that knows how those facts depend on one another, and it is the reason a connected view is hard to assemble by hand.
26 August 2026 · 9 min read
Rather talk it through?
We learn most from real operational problems. Tell us where earlier intelligence would make a difference, or call {PHONE}.