An emergent FinTech startup is developing a mobile banking application which anticipates sporadic and significant usage peaks, primarily during monthly payroll periods. They need to implement a feature that processes various customer transactions and runs complex computations on demand. The startup aims to maintain infrastructure management to a minimum while ensuring costs remain aligned with their actual consumption levels. Which option is the MOST suitable for the dynamic transaction processing component of their application?
Use AWS Batch to manage transaction processing jobs, taking advantage of its ability to efficiently run batch computing workloads across a full EC2 instance fleet.
Utilize AWS Lambda functions triggered by the application, ensuring on-demand scaling and billing for compute time without server management.
Configure an Amazon SQS queue to decouple incoming transactions and process them using an Auto Scaling group of EC2 instances based on queue length.
Deploy the computational logic to a managed Kubernetes service using Amazon EKS, leveraging Kubernetes Horizontal Pod Autoscaler to scale based on demand.
Implement a server fleet using Amazon EC2 with Scheduled Scaling to handle expected peak periods based on predictable payroll cycles.