Great write-up, Matias! Looks like Skillshare DevOps has made some great improvements in the past few months, which is exciting to see. In particular, I remember cluster auto-scaling as being something that we always wanted to get working, but we didn’t have an easy way for the cluster to report up to the AWS autoscaling group that there wasn’t enough room to schedule all the pods. The cluster autoscaler documentation you linked to only mentioned GCE/GKE. Does it also work out of the box for AWS?