// build systems that survive production
In-depth, production-tested engineering — deep dives on backend, cloud, AI, DevOps, databases, and system design, for developers who build systems that have to survive real traffic.
Networks fail, clients retry, and duplicate requests happen. Here's how to design write endpoints that produce the same result no matter how many times they're called.
Why JSON logs with stable fields beat grep-and-pray when production is on fire.
Rolling out safely with flags — and remembering to delete them.
Postgres until you can prove you need something exotic.
Queues, load shedding, and what happens when producers outrun consumers.
Where else exactly-once thinking saves you from duplicate side effects.
Practical levers for trimming that first-request latency.
Turning Postgres query plans into actionable fixes.
Handling SIGTERM so deploys don't drop in-flight requests.
APIs, services, queues, and the unglamorous plumbing that keeps systems running.
Deployments, infrastructure, containers, and running software you don't babysit.
LLMs, embeddings, vector stores, and shipping AI features that actually work.
Trade-offs, scaling, consistency, and reasoning about distributed systems.
CI/CD, observability, reliability, and the path from commit to production.
Query plans, indexes, transactions, and getting the most out of your data layer.
Production-grade engineering writing in your inbox. No spam, unsubscribe anytime.