As a developer, I am requested to optimise only:

* Bootstrap the runtime
* Code run
* Use arm64 architecture

Serverless engineering means we need to take care of many moving parts, software, architectures, security and so on.

The best practise developing with AWS Lambda functions are:

* The Lambda package should be as small as possible
* Initialise my classes, SDK clients and database connections outside of the function handler
* Cache static assets locally in the /tmp directory.
* Doing all this will save execution time and cost for subsequent invocations (Warm start).

On top of these basic rules, maybe it is good to mention:

* Avoiding fat lambda function
* Executing code in parallel when possible

When we are talking about serverless, speed is essential. I am not talking about spending hours of development time saving a few ms, but I am talking about the relation between speed and cost without moving a finger. Speed in a world hyperconnected where everything must be available in real-time is essential, and for few that use this concept at scale, it will result in a cheaper cloud cost.

Rust is the key to unlocking the performance of the next generation of serverless applications. However, thanks to the Single Responsibility Function that reduces the code complexity, Rust has become just a tool, another language syntax that allows me to increase the execution speed more than 9 times. Speed is money.