Nowadays the smart phone market is clearly growing due to the new type of functionalities that devices have and the role that they play in everyday life. Their utility and benefits rely on the applications that can be installed on the device (the so-called apps). Cloud computing is a way to enhance the world of application by providing disk space and freeing the user of the local storage needs, this way providing cheaper storage, wider acces-sibility and greater speed for business. In this paper we introduce various aspects of computing and we stress the importance of obtaining cloud maturity by using machine learn-ing for automating configurations of software applications deployed on cloud nodes using the open source application ERP5 and SlapOS, an open source operating system for Decentral-ized Cloud Computing.Keywords: Cloud Computing, Mobile Computing, SlapOS, Machine Learning, ERP5(ProQuest: ... denotes formulae omitted.)1 IntroductionIn our opinion the triggers of today's and future IT market are devices, cloud and social networks. According to Gartner's analysis, mobile experience is overshadow-ing the desktop experience [1]. The client-server architecture is leftbehind and cloud computing put together with devices draw the modern cloud-client architecture, where the cloud client keeps the usual pro-gramming platform and programming lan-guage model. From the user experience per-spective, cloud combines the benefits of ac-cessing services and having available storage settings. One of the most valuable features of cloud computing is represented by mobility.It is a fact that smartphones are becoming ubiquitous as, in 2012; the smart phone shipment on a global level grew 43 percent reaching 700 million units, according to Re-search firm Strategy Analytics [2]. Also, ac-cording to Credit Suisse, predictions for 2014 estimate one billion devices sold solely in that year and by 2016 there will be at least ten billion smartphones connected to the In-ternet via apps [3].Also besides this new feature of Cloud Com-puting it is important to reach Cloud Maturi-ty. That is why this article describes the methodology to design and implement an open source test platform for machine learn-ing and Enterprise Resource Planning (ERP). In addition, we present an approach to test different machine learning solutions and an example of results obtained by using ERP5. The conclusion summarizes the contribu-tions.2 Mobile ApplicationsIn this chapter we analyze the evolution of development which now raises the expectations from design and development points of view. The operating systems used until now brought their advantages. Among these, there are Android, Blackberry, iOS, Symbian, WebOS, Windows Phone, Bada, Tizen, but the features that meet higher standards are offered today as for example, by Firefox OS, an open source operating sys-tem. This operating system is based on a HTML5 web browser and new APIs from Mozilla. HTML5 represents a new frame-work that enriches the user web experience across devices. It represents an intelli-gent usage of less code to optimize web ap-plications. Users expect to access the pub-lished content through a wide range of plat-forms with a minimum engaged complexity [4].The evolution from the usual con-sumer applications to more complex enter-prise applications goes toward a single tool that matches all the applications. An issue raised by the content of applications is represented by the request to develop differ-ent renditions based on various features of the devices. A common approach would be to have applications built upon open stand-ards, but there is reluctance towards adopting open applications. Native applications offer a package of content and services and they permanently verify what it occurs on their own platforms. These applications put a bar-rier on what HTML5 web applications are able to accomplish from the technical point of view and business motives. …