A data analytics firm runs large-scale batch jobs that process high volumes of data using Amazon EC2 instances. These jobs are stateless and fault-tolerant, meaning they can be interrupted and resumed without data loss or impact. Each batch run typically lasts around 60 minutes and can be triggered at unpredictable times. The company is looking for a scalable and cost-effective solution to handle these workloads while keeping infrastructure costs low. What should a solutions architect recommend?