Rapid Development of Service-based Cloud Applications: The Case of the Cloud Application Platforms

Introduction

In recent years two major technological trends have emerged and are able to drive the evolution of service oriented computing, namely cloud computing and Big Data. The rise and proliferation of cloud computing (Armbrust et al., 2010) and cloud platforms in specific (Cusumano, 2010), has the potential to change the way cloud based software applications are developed, distributed and consumed. Cloud platforms popularity stems from their potential to speed up and simplify the development, deployment and maintenance of cloud based software applications. Nevertheless, there is a large heterogeneity in the platforms offerings (Gonidis, Paraskakis, Simons, & Kourtesis, 2013) which can be classified into three clusters. On one cluster application development time is drastically decreased with the use of bespoke visual tools and graphical environments at the expense of a restricted application scope which is usually limited to customer relationship management (CRM) and office solutions. At the other end of the spectrum platforms offer basic development and deployment capabilities such as application servers and databases. The intermediate cluster consists of cloud platforms, which offer additional functionality via the provisioning of, what the authors call, platform basic services (e.g. mail service, billing service, messaging service etc.). A platform basic service can be considered as a piece of software, which provides certain functionality and can be reused by multiple users. It is typically provisioned via a web API. The platforms offering such services are also referred to as cloud application platforms (Kourtesis, Bratanis, Bibikas, & Paraskakis, 2012). The rise of the cloud application platforms has the potential to lead to a paradigm shift of software development where the platform basic services act as the building blocks for the creation of service-based cloud applications.

Almost in parallel with the emergence of cloud computing, data volumes started skyrocketing leading to what is commonly referred nowadays as Big Data (Jacobs, 2009). The term Big Data has been initially attributed with certain features (McAfee, & Brynjolfsson, 2012) such as: a) Volume, b) Velocity and c) Variety. Volume refers to the large amount of data which are generated each day. Velocity dictates the need for analysis of the rapidly collected data as those are becoming outdated quickly. Variety denotes the diverse forms and formats in which data are collected and stored. In addition to these features, Demchenko et al. (2013) propose a wider definition of Big Data by adding two extra characteristics: d) Veracity and e) Value. Veracity refers to the amount of uncertainty that the collected data contain and to the extend they can be considered trustworthy. Value denotes the added-value that data can bring to the predictive analysis. Therefore, what once used to be a technical problem due to limited storage and processing capacity now it is being transformed into a business opportunity (The Data Warehouse Institute, 2011).The collection and analysis of Big Data can lead to the extraction of meaningful information and can subsequently drive important decisions (McAfee, & Brynjolfsson, 2012).