Automatically adjusting the number of servers based on demand. Like a restaurant that opens more tables during busy hours and closes them when it's quiet.
During Black Friday, Auto Scaling automatically adds more EC2 instances to handle increased website traffic, then removes them when traffic returns to normal.
All four platforms automatically add or remove compute instances based on metrics (like CPU), schedules, or load balancer demand. AWS uses Auto Scaling Groups for EC2; Azure scales VMSS; GCP scales instances in a Managed Instance Group; OCI scales instances in an Instance Pool. Each integrates with load balancers and monitoring to scale out/in while maintaining availability.