In this chapter, we have seen two different philosophies relating to performance improvement. At the beginning, we tried to reduce the bandwidth used by our clients by caching data and using as few connections to our server as possible.

In the second part, though, we began to do something more advanced by allowing searches to be run in parallel and each client to remain in sync with a persistent connection to the server through web sockets. This will allow clients to receive updates in real time, and our application will feel more reactive but consume more threads.

I strongly encourage you to polish the result before we move on to the next chapter and deploy our application for good!