Blog
Practical analysis of software failures, test automation, AI Assurance, and building dependable products, alongside essays and reflections.
Filter by tag
A retrospective on the July 2024 CrowdStrike incident: testing configuration contracts, runtime behaviour, staged delivery, and recovery as one release system.
A practical AI Assurance analysis of Gemini’s 2024 image-generation failures: task-specific acceptance criteria, evaluation sets, refusals, and regression evidence.
A retrospective on testing distributed workflows during network partitions: consistency, queues, retries, degraded behaviour, and recovery evidence.
How a rule can return the right answer and still fail the product: performance test design, adversarial inputs, budgets, and safe rollout.
A practical recovery-testing approach inspired by GitLab’s 2017 database incident: restoration, data integrity, business workflows, and observable recovery targets.
What Knight Capital’s 2012 trading incident teaches about mixed-version deployments, business invariants, rollback, and operational stop controls.
A retrospective on the July 2024 CrowdStrike incident: testing configuration contracts, runtime behaviour, staged delivery, and recovery as one release system.
A practical AI Assurance analysis of Gemini’s 2024 image-generation failures: task-specific acceptance criteria, evaluation sets, refusals, and regression evidence.
A retrospective on testing distributed workflows during network partitions: consistency, queues, retries, degraded behaviour, and recovery evidence.
How a rule can return the right answer and still fail the product: performance test design, adversarial inputs, budgets, and safe rollout.
A practical recovery-testing approach inspired by GitLab’s 2017 database incident: restoration, data integrity, business workflows, and observable recovery targets.
What Knight Capital’s 2012 trading incident teaches about mixed-version deployments, business invariants, rollback, and operational stop controls.
Output is now everywhere. Code, designs, features, and prototypes are easier to produce than ever. The real engineering advantage is judgement.
Accessibility is a business decision that determines whether real customers can buy, book, subscribe, compare, apply, and trust your website.
A clear technical breakdown of CORS, what triggers it, what causes preflight, and the common and uncommon cases that make browser requests fail.
Full-screen blocking preloaders are rarely an intentional UX decision, but they are usually an architectural side effect. Here is why they hurt marketing pages and what to do instead.
SEO can drive visibility, but without business clarity, trust, and operational readiness, traffic alone will not convert.
The beginner-friendly handoff contract that scales (organized files, states, components, and responsive behavior) so engineering ships without guesswork.
How I built a repeatable CI/CD pipeline to deploy Astro projects to cPanel shared hosting using GitHub Actions, eliminating FTP, avoiding platform lock-in, and navigating real-world deployment failures.
Here is the measurement stack I use, when I use each tool, and what I look for.
I’ve been writing for years and hiding it in forgotten folders. This is me breaking the cycle (publishing anyway) and building a home for my thoughts, notes, and poems.