A high-performance API for smart content-driven mobile applications

Abstract: Nowadays, smartphones are one of the main mediums to transfer content to clients and they provide a good interface for the delivery of digital therapeutics (DTx). Application Programming Interfaces (APIs) are keystones for any service-based mobile application to manage the data transfer between the clients and the server(s). At the time of writing, REST and GraphQL are arguably the most popular API types. These APIs provide simplicity, efficiency, and scalability for data-driven applications. However, we can question the validity of these qualities when it comes to content-driven applications such as DTx apps for example. The objective of this project is to decide whether common data-driven API types are suitable for products where the content is not created by the users, but by content creators, or if a new API type could be developed which is more suitable for this use case. Useful insights are presented on how standard APIs should be adapted to address this problem. Moreover, this document presents a new prototype of API that is supposed to help content creators with content creation while providing high performance. Additionally, an advanced test platform is introduced in order to help the comparison of the implementation in terms of performance. This thesis shows that GraphQL is not a suitable method for the initial problem. It also shows that REST is a good standard adapted for scalability in the present scope. The prototype displayed guarantees great performances by means of preloading, thus, suffers high concurrency in terms of performances when scaled up. However, potential solutions are given in order to counter this.