If you're not familiar with the concept, serverless services like AWS Lambda and DynamoDB On-Demand allow you to pay only for the resources you consume rather than maintaining always-on infrastructure.

This can be a game-changer for workloads with variable or unpredictable traffic, such as web servers, async workers, and even databases.

By shifting to a serverless model where appropriate, you can significantly reduce your overall spend and improve your application's scalability.

How do you determine if a service would be better run as serverless? Generally if the service is sporadically run then it is a good candidate for serverless. If it is constantly getting hammered with traffic or the few extra milliseconds it would take for a serverless service to boot up each time it is spun up would have a negative impact on your application then you probably want to go with an always on option.

What questions do you have about adopting serverless in your own environment?