The onboarding of a new strategic client on the Paidy platform brought with it stricter latency requirements (in order to meet the agreed-upon Service Level Agreement). The backend application built on top of Akka Cluster with cluster sharding did not deliver consistent latency in redeployment and failover scenarios. I helped the engineering team find and understand the root cause of the latency fluctuations and guided them throughout the process of implementing a solution.

Topics & tools: Performance Engineering, Akka Cluster, Akka Cluster Sharding, AWS ECS