Blog
Notes from
the build
Practical, opinionated writing on engineering durable software: modernization, platforms, reliability and how to choose a partner you will not regret.
Rocket Software Buys Geniez AI to Put Agents on the Mainframe
Rocket Software acquired Geniez AI on October 13 to accelerate agentic AI for mainframe modernization. Here is what the $100M deal signals about legacy core systems.
When to Rewrite vs Refactor a Legacy System
A decision framework for choosing between refactoring and rewriting, based on architecture fit, risk, and audit needs.
Google Cloud Storage's 12-Hour Outage: What the Status Page Delay Actually Cost
A traffic optimization change knocked out GCS writes in us-central1 for 12.5 hours, and Google's own status page lagged seven hours behind reality.
Technical Debt Has an Interest Rate, Here Is How to Calculate It
Technical debt compounds like a loan. Here is the formula to calculate its interest rate and decide when paying it down beats letting it ride.
Schneider Electric Buys PTC: What a $22.6B Acquisition Means for Your Stack
Schneider Electric is paying $22.6 billion for PTC. If Windchill, Onshape, or ThingWorx sit under your product data, here is what an acquisition like this actually changes.
The Real Cost of Vendor Lock-In (And How to Avoid It)
Vendor lock-in rarely shows up as a line item until you try to leave. Here is what it actually costs and how to design around it.
AWS Proton's End of Support: A Lesson in Platform Vendor Risk
AWS is ending support for Proton on October 7, 2026. The infrastructure survives, the orchestration layer does not. Here is what that means for build vs buy.
Designing for Failure: Chaos Engineering Basics for Non-Unicorns
You don't need a platform team or Netflix's scale to start chaos engineering, just a hypothesis, a blast radius, and the discipline to run the experiment anyway.
Azure's 18-Region Gateway Outage: What Engineering Leaders Should Take From It
A gateway management change collided with routine OS servicing and knocked out Azure networking across multiple regions. Here is what it means for teams betting on a single cloud.
How to Choose a Custom Software Partner in a Regulated Industry
In regulated industries, the wrong software partner does not just ship late, they ship risk you inherit for years.
Modernizing Legacy Systems Without a Big-Bang Rewrite
The full rewrite is the most seductive and most dangerous idea in software, and there is a calmer way through.
Platform Engineering for Mid-Size Companies: When It Pays Off
An internal developer platform is a force multiplier or an expensive distraction, and the deciding factor is usually your size.
Building Software That Survives an Industry Downturn
Downturns do not kill good software, they kill software that was too expensive, too fragile, or too locked in to survive scrutiny.