Decision Journal
Structured thinking, in writing
Notes on enterprise AI adoption, customer discovery, operational decision-making, and what it actually takes to turn a proof of concept into a deployed system.
2 June 2026 · 2 min read
Enterprise AI Adoption Starts With Workflows, Not Models
Most AI adoption programs fail before the model is ever evaluated, because the workflow underneath it was never mapped.
14 May 2026 · 2 min read
Customer Discovery Is a Technical Skill, Not a Soft One
Engineers who treat discovery as 'the PM's job' end up building technically correct solutions to the wrong problem.
22 April 2026 · 2 min read
Data Fragmentation Is the Default State, Not the Exception
Plan for fifteen disconnected systems from day one. The organisations with clean, unified data are the exception, not the baseline.
30 March 2026 · 2 min read
From PoC to Production: the Gap Nobody Scopes
A proof of concept and a production system solve different problems. Most projects only budget for the first one.
5 March 2026 · 2 min read
Designing AI Workflows: Keep Humans at the Decision, Not the Task
The question isn't whether a human is 'in the loop.' It's which loop they're in: the task loop or the decision loop.
18 February 2026 · 2 min read
What AI Governance Actually Means in Practice
Governance isn't a policy document. It's the set of answers you can give when something goes wrong at 2am.
20 January 2026 · 2 min read
Business Outcomes Over Technology: Lessons From Enterprise Deployments
The best technical decision and the best business decision are usually the same decision, viewed from different angles.