~/index
Ceiba has a CloudWatch alarm that stops the production instance at $70 to keep the bill under $80. It had never actually run. This is what happened when I made it run, three times, and the instrumentation gap it exposed.
Ceiba left the homelab for AWS. Notes on the architecture, the $34-a-month budget that shaped every decision, a CI/CD migration that turned out not to be flaky after all, and what I'd already do differently.
How I’m using a single-node platform to test Ceiba as a real workload, validating runtime, data, backups, and scaling boundaries before moving toward multi-node or cloud infrastructure.
Notes on API access, product boundaries, and building a useful first version before chasing broader architecture.
How I designed storage and operational boundaries on a single-node homelab to reduce blast radius, improve recoverability, and build platform discipline before workloads.
This site isn’t only about code — it’s about systems, constraints, people, and the craft of building things that can be operated and trusted.
A practical story of moving from brittle tenant fetches to a clean, type-safe SSR flow with Orval, Next.js App Router, and ASP.NET Core.
How we made tenant branding feel custom without rebuilding CSS: SSR-injected variables, Tailwind utilities, and a theming model that stays maintainable.
How I wire SEO in Next.js App Router: metadata, JSON-LD, sitemaps, and performance—explained in plain language, with a setup that stays maintainable.
