S3zipper can run on AWS Fargate, so you do not have to manage servers. The task definitions are public:
github.com/elgonlabs/s3zipper-deploy
A typical setup uses two services from the same image: a web/API task behind an application load balancer, and a worker task that consumes the job queue. Provide the database connection, the queue (AMQP or Redis), and the API secret as environment variables, and make sure the task role can write to the buckets you zip into.
This page is a short pointer; for a discussion of the moving parts, open an issue on the repository or contact us.