From 1k to 100k Users: What We Learned Scaling a Fintech Platform

Scaling a fintech product is not only about handling more requests per second. It is about audit trails, reconciliation, idempotent payments, and knowing exactly what happened when something goes wrong at 2 a.m.

Early on, we invested in observability: structured logging, tracing across services, and alerts that fire on business metrics, not only CPU graphs. That paid off when traffic spiked after a partner integration went live.

What mattered less than we expected: rewriting the API in a trendier language. What mattered more: database connection pooling, read replicas for reporting, and a clear separation between transactional workloads and analytics.

The near-miss? A batch job that touched the same rows as live transactions. We fixed it with explicit locking boundaries and moved heavy reporting to a replica. Lesson: scale the boring parts first.