Default values of runtime parameters are rarely adequate for each application and may actually have negative impact on application performance.
Tuning runtime may have dramatic impacts on performance and resilience, without requiring any code changes. Moreover, it can also hugely impact cost efficiency by reducing the demand of container or cloud resources.
However, the number of available options (e.g. JVM has 800+ tunable flags) makes this an impossible task to be performed at each application release.
Akamas leverages AI to recommend (and apply) runtime configurations that make applications deliver the best quality of service and cost efficiency against any custom-defined SLOs.