Have you ever picked up a project that became so complex you felt like a 'fresher' all over again?

One such challenging project I delivered was a crypto exchange engine, where the objective was to deliver 2K+ TPS in a vertically scaled system to avoid conflicts and minimize latency.

When traditional CRUD APIs and SQL architectures failed, we developed an in-memory engine in C, backed by the resiliency of MySQL and Kafka for the backup and restoration of the in-memory state using sequential logs.

The result 10K+ TPS per CPU Core with a proven scalability to 1M+ TPS

Just added the full case study – high-performance-cryptocurrency-trading-engine