So let's break it down into the two different scenarios.
1. Reloading Applications
The management of the platform, user access control, provisioning of applications and data reload are done through what the refer to as the Qlik Repository Service API. This is a REST based api so any programming language that is able to execute HTTP requests and parse JSON will be able to integrate with it.
2. Embedding Analytics
3. (which you didn't ask for ) Communicating with the Qlik Engine
If you do not wish to leverage any Qliks visual components but instead leverage our on-demand, in-memory calculation engine you would leverage the Engine API. The Engine API is a protocol that defines how the communication between clients and the Engine should take place. So as long as your programming language supports WebSockets and JSON then you can initiate communication with the Qlik Engine, if you want to perform batch, server-side tasks or building your own client.
I hope that made things a little bit easier to understand
If you're looking into building a function for preloading apps to memory, you might want to look at the "App preloader" example at the following page:
There is also a more elaborate version of the example on branch: