Keynote talk 2: How to build an industrial R&D center in Vietnam: A case study

Summary form only given. To be competitive in today's market, the IT industry faces many challenges in the development and maintenance of enterprise information systems. Engineering these largescaled systems efficiently requires making decisions about a number of issues. In addition, customers expectations imply continuous software delivery in predictable quality. The operation such systems demands for transparency of the software in regard to lifecycle, change and incident management as well as cost efficiency. Addressing these challenges, we learned how to benefit from traditional industries. Contrary to the fact that the IT business calls itself gladly an industry, the industrialization of software engineering in most cases moves on a rather modest level. Industrialization means not only to build a solution or product on top of managed and well-defined processes, but also to have access to structured information about the current conditions of manufacturing at any time. Comparably with test series and assembly lines of the automobile industry, each individual component and each step from the beginning of manufacturing up to the final product should be equipped with measuring points for quality and progress. Even one step further the product itself, after it has left the factory, should be able to continuously provide analytic data for diagnostic reasons. Information is automatically collected and builds the basic essentials for process control, optimization and continuous improvement of the software engineering process. This presentation shows by means of a practical experience report how AdNovum managed to build its software engineering based on a well-balanced system of processes, continuous measurement and control — as well as a healthy portion of pragmatism. We implemented an efficient and predictable software delivery pipeline