Practical thinking on cloud architecture, reliability engineering, and AI integration — written from 27 years of hands-on experience, not analyst reports. We publish when we have something worth saying. The first essays are in progress and will be available shortly.
Every major technology transition follows the same pattern — economics shift, standards form, talent follows, tooling matures, and adoption becomes inevitable. AI is no different. What the mainframe era, the web era, and the cloud era teach us about navigating what comes next — and why the organizations that win are rarely the first movers.
Notify me when publishedIt is rarely a technology problem. It is almost always a transition problem — organizations adopting new platforms without understanding the forces driving the shift.
When the cost curve shifts, adoption becomes inevitable. The real strategic question is not whether to adopt AI — it is when, and what your infrastructure needs to look like when you do.
Many organizations migrated to cloud infrastructure without migrating to cloud thinking. The result is a generation of estates that carry the cost of cloud without the flexibility it was designed to deliver.
The principles that kept large-scale enterprise systems running under strict SLAs in the 2000s are more relevant to modern cloud-native teams than most engineers realize.
Mainframes still process the majority of the world's financial transactions. Understanding why tells you something important about how technology transitions actually work — and what AI adoption will really look like.
Most organizations are AI-experimenting, not AI-native. The infrastructure decisions made in the next 24 months will determine which organizations are able to move fast in this era — and which will spend years catching up.
We write when we have something worth saying. Subscribe and we will notify you when new essays, frameworks, and resources are published — no noise, no cadence for its own sake.