For our backend, we've chosen to go with a RESTful interface using API Gateway, Lambda, and DynamoDB. We're already familiar with deploying these services from earlier in Chapter 4, Adding Amazon API Gateway, and Chapter 5, Leveraging AWS Services. What we also need here is to add Route 53 so that we have a public DNS record that matches our domain.

We could also add some identity functionality with Cognito—if we had data related specifically to a user, we might want to protect that endpoint to validate the identity of the requestor. This would be set up as an authorizer for API Gateway to match a valid JSON Web Token (JWT) with a user, and to implement login and sign up flow.

For DynamoDB, if our request load was significant and consistently hitting a particular table or tables, we could investigate using DynamoDB Accelerator (DAX). This is a caching layer for DynamoDB that can reduce the latency of our queries—in some cases, down to the microsecond. 

And that's all we need to get going. The components here will allow us how to scale to millions of users without any significant input from administrators. There are also now more options than ever to represent these components as code, in a single framework. We'll have a look at one option from a high level.