It is a bit like Kubernetes, you end up needing 3 chiefs and 12 workers to run something that a single one was doing before. Then you realize that anyway you can't reuse that process for something new you are trying to do and need another team of 3+12...