// build log
Engineering Notes
Things I've learned building distributed systems, debugging production failures, and figuring out what CIS Benchmark controls actually mean in practice.
Built CloudFlow using choreography — each service reacts to events independently. No central brain. Here's what I learned about failure modes, compensating transactions, and idempotency under load at 1,100+ req/min.
After building a CSPM tool that scans against CIS controls, here's what I found: most teams fail the same 4 IAM checks. The fixes are straightforward. The organisational politics around implementing them are not.
Remote state, workspaces, state locking, import vs refactor — a write-up of every mistake I made setting up Terraform for a multi-environment AWS setup from scratch. Some of these cost hours.
OOMKilled pods, throttled CPUs, and why setting limits too tight is almost as bad as not setting them at all.
They look the same. They're not. A guide to when to use which — and what breaks when you pick wrong.
How writing a formal spec for AgriFuture India changed the way I think about race conditions in every system I build.