An application stores data in a DynamoDB table using provisioned capacity. The workload has predictable scheduled peaks (an early Monday morning surge) and very low weekend usage. The company needs consistent performance during peaks at the lowest cost. Which solution is most cost-effective?

The most cost-effective solution is to use AWS Application Auto Scaling to schedule capacity adjustments. This approach directly addresses the predictable nature of the workload. By scheduling higher provisioned capacity for known peak times (like Monday mornings) and lower capacity for off-peak periods (like weekends), the application maintains consistent performance when needed without incurring unnecessary costs during idle times. Increasing provisioned capacity to the maximum peak load permanently would ensure performance but would be expensive due to over-provisioning during off-peak hours. Dividing the table and spreading queries is complex and doesn't inherently solve the cost issue for fluctuating demand. Changing to on-demand capacity would handle scaling automatically but is generally more expensive than provisioned capacity for predictable workloads, as on-demand pricing is higher per unit of capacity.