Author: Martin Morcate Trujillo

A global investment bank and Confluent used Apache Kafka to deliver sub-5ms p99 end-to-end latency with strict durability. Through disciplined architecture, monitoring, and tuning, they scaled from 100k to 1.6M msgs/s (<5KB), preserving order and transparent tail latency.