Case Study: Sisal achieves 58% production cost reduction with Akamas

How a global gaming leader leveraged Akamas' AI-powered, full-stack, live optimization platform to safely cut its cloud costs

Sisal, a leading gaming operator headquartered in Milan, needed to optimize the performance and cost of its Kubernetes-based, Java microservices supporting the Piattaforma Gestione Terminali (PGT) application used across thousands of point-of-sale locations. The company wanted to improve throughput and response times in production without risking downtime or degrading the user experience, while also avoiding unnecessary over-provisioning in its cloud environment. Akamas’ AI-powered optimization platform was chosen to help address this live, full-stack tuning challenge.

Akamas integrated with Sisal’s observability stack, including Dynatrace, and performed live optimization in production across Kubernetes and JVM settings, using safety policies to make gradual, low-risk changes. By continuously adjusting CPU, memory, heap size, and garbage collection parameters while preserving service-level goals, Akamas reduced production costs by 58% in less than two weeks, including a 40% reduction in memory consumption and a 68% reduction in CPU usage.

Sisal

Carlo Monaco

IT Operation Director