# Profile and measure the actual bottleneck before attempting any architectural change to address it