How would you set up Kubernetes clusters and auto-scaling groups to handle computational load for a new analytics tool at Dun & Bradstreet?
