Information system development process improvement in the case of banking system

The thesis is based on the case of developing banking information system in the bank where I am working. The banking information system is a result of our own knowledge and is based on the existing development process that was set in 2005. The process is based on the waterfall and spiral model, the classical principles of project management and UML modeling techniques. All phases of the existing development process are presented in the first part of the thesis. The participant roles, tecniques and key products generated during the process form the basis of the methodology of information system development in the bank. The current financial situation, software development trends lead us to change the way we work if we want to maintain the workload and improve the quality of the banking system. Therefore, the second part of the paper presents methodologies that enable us to achieve those goals. Among presented methodologies the most important are lean software development and agile methodologies like Scrum, extreme programming (XP), agile unified process (AUP), adaptive software development (ASD), disciplined agile delivery (DAD). Above all, lean principles are those who bring great benefits to entire organization, but require a change in mindset at all levels. As described in section 4 this means that the implementation is more of a long-term issue. The main objective of the thesis is to apply a modern agile and lean development approaches to the existing process of developing banking information system. The third part describes proposed new development process and suggests ways to apply the process to the entire IT in the bank. The success of the new development process depends on cooperation and changing the mindset of all involved in the software development process. The key factors of new development process are: preparation and maintenance of product backlog, creating of heterogeneous development teams and short, fixed-time development cycles. In the thesis we have set ourselves four thesis, which can all be confirmed as being the foundation of modern software development approach. They say that we have to minimize the number of phases in the new development lifecycle, use automated testing methods and maintain unique prioritized requirements backlog