The Java 8 web application currently running uses embedded Jetty and needs to serve hundreds of thousands of requests per second. Which it does, thanks to AWS auto-scaling. However, our intuitive feeling is that there are a number of things that can be optimized to make this use much fewer servers than currently.
What is most needed is understanding of how to profile software applications, with focus on JVM, associated tools and agents, and of course understanding of OS (Linux in this case) and its implications (network, filesystem, threads and processes) on this. Although you will be spotted on this part by the IT folks.