The database workload has high variability and unpredictability, which makes fixed-capacity database solutions inefficient and costly. The company also requires MySQL compatibility without schema or application changes.

Amazon Aurora Serverless (MySQL-compatible) is specifically designed for this scenario. Option C automatically scales database capacity up or down based on workload demand, making it highly cost-effective for sporadic usage patterns. During peak periods, Aurora Serverless scales seamlessly to handle increased load, and during idle periods, capacity scales down, significantly reducing cost.

Option A (DynamoDB) is not suitable because it is a NoSQL service and would require significant application and schema changes. Option B (RDS for MySQL) uses fixed instance sizes and would require overprovisioning to handle peak loads, increasing cost. Option D introduces high operational overhead and does not provide automatic database scaling.

Therefore, C best meets the requirements by providing MySQL compatibility, serverless scaling, and cost optimization for variable workloads.