A company's media-sharing platform encounters fluctuating levels of user engagement, leading to sporadic performance bottlenecks and the overprovisioning of resources during quieter periods. Which service should be implemented to automatically adjust the quantity of compute resources to align with the real-time workload demands and achieve cost savings?
Deploy a serverless architecture for all the backend processes
Allocate a constant quantity of larger compute instances
An elasticity provider that adjusts capacity in real-time
Manually scale the compute instances based on projected usage