Release Notes

We are excited to introduce Akamas Version 3.2.2, this minor release focus on usability improvements to enhance your optimization experience. As always, we extend our heartfelt thanks to our community of users for their invaluable feedback, which continues to drive the growth and improvement of Akamas. Let's take a closer look at the enhancements in this version!

Usability Improvements

Users can now easily differentiate telemetry instances by providing custom names. The logs of the execution of telemetry instances have been improved and split by instance name to ease the discovery and resolution of configuration issues, particularly in setups with multiple telemetry providers.

When deleting a system associated with a running study, users now receive a confirmation prompt to make informed decisions and avoid accidental deletions.

Take the most out of your data. Even if the computation of the score fails (e.g. due to a missing key metric), Akamas now computes trial stability windows and aggregated metrics in order to quickly identify the performance of the configuration with respect to other KPIs and constraints.

Deploying Akamas on Kubernetes is now a one-liner process. Additionally, we've improved the startup, liveness, and readiness probes for a seamless experience.

We value your input and hope these usability improvements empower you to optimize your systems effortlessly. Thank you for choosing Akamas, and we look forward to your feedback as we strive to make Akamas even better in the future!