Principles that hold whether the work is a website, an automation, a Cognitive AI deployment or a custom system.
Start from the result. The software serves the result, not the other way around. Do not build abstractions before understanding the operational reality.
Approvals, permissions and evidence are built in. Accountability is a structural requirement, not a feature toggled on later.
When the world changes, the system responds. Rules are visible and editable by the people doing the work.
You can trace what happened, and why. Black box decision making is rejected in favour of clear, auditable logs.
Systems remain useful and understandable after handover. Complexity is reduced so maintenance is predictable.