Engineering articles
8 articles in this category

How we built a production email system handling 18 transactional email types — from auth verification to billing receipts — with SES domain verification, template rendering, and delivery monitoring.

How we built a 4-tier WebP thumbnail pipeline that generates responsive variants on upload, cutting blog image bandwidth by 93-99% — from 550KB originals to 5KB admin thumbnails.

How we built a complete CI/CD pipeline with Docker multi-stage builds, Kubernetes deployments, real-time log streaming, and GitHub Actions webhooks — all managed through a single dashboard.

How we designed a storage abstraction that supports both local filesystem and S3 — with environment-driven switching, URL derivation, and zero application code changes when migrating between providers.

Why local disk storage doesn't scale for a SaaS platform, how we integrated S3 with CloudFront on cdn.boottify.com, and the storage abstraction that makes provider switching seamless.

How we identified and eliminated all 7 files over 500 lines, created shared hooks and components, and brought our entire codebase under control without breaking anything in production.

Major outages, aggressive pricing changes, and AI-driven workload diversity are pushing teams toward multi-cloud. Here's why the single-cloud era is ending and what a realistic multi-cloud strategy looks like.

GPU scheduling, bursty inference traffic, and multi-terabyte model weights are breaking traditional SRE assumptions. Here's how Kubernetes and the SRE discipline are adapting to the AI workload era.