Store as much analytics and game activity data as financially feasible today so it can be used to train machine learning models to predict user behavior in the future.

Begin packaging their game backend artifacts in container images and running them on Kubernetes Engine to improve the availability to scale up or down based on game activity.

Set up a CI/CD pipeline using Jenkins and Spinnaker to automate canary deployments and improve development velocity.

Adopt a schema versioning tool to reduce downtime when adding new game features that require storing additional player data in the database.

Implement a weekly rolling maintenance process for the Linux virtual machines so they can apply critical kernel patches and package updates and reduce the risk of 0-day vulnerabilities.