Essays on enterprise architecture, integrations, data and applied AI, written from practice. RSS
[Essay · How a data tool gets adopted 4/4 · 6 min · Jul 14, 2026]
The real binary isn't asking permission or forgiveness. It's building evidence before consensus exists. When that play works, when it becomes governance debt, and how to move from internal hacker to institutional architect.
[Essay · How a data tool gets adopted 3/4 · 8 min · Jul 10, 2026]
Through the lens of social control and criminology: how conduct that breaks formal procedure can be reframed as innovation when it produces utility, buy-in, and legitimacy. Organizations don't judge only the rule, but the outcome, the context, and who benefits.
[Essay · How a data tool gets adopted 2/4 · 7 min · Jul 7, 2026]
The persuasion stack that drove adoption of a data tool wasn't narrative: it was coded and enforced server-side. Atomic attempt limits, exam windows, idempotent certificates, and a notification pipeline as a delta job.
[Essay · How a data tool gets adopted 1/4 · 6 min · Jul 3, 2026]
Adopting an internal tool isn't a feature problem. It's a trust and status problem. How an internal business school closed a data unit's credibility gap and turned the tool into a reference.
[Essay · 7 min · Jun 23, 2026]
How property-based testing surfaced a latent bug in an ERP's tax engine, a bug no example test was ever going to touch, and how we wired it into the workflow across two stacks: PHP with eris, and TypeScript and Python in the data lake.