In praise of small systems.
Small does not mean simplistic; it means the whole thing still fits in your head.
A small system can be understood end to end. You can see where data enters, how it changes, and what comes out. That visibility makes maintenance less mysterious and decisions more reversible.
Small systems still solve serious problems. Their advantage is that complexity is introduced in response to evidence, instead of being prepaid for a future that may never arrive.
The right amount of architecture is enough structure to make the next change unsurprising.
This is not an argument against scale. It is an argument for earning scale deliberately: measure the pressure, locate the constraint, and add sophistication exactly where it changes the outcome.