Undulate: A framework for data-driven software engineering enabling soft computing

Especially web-facing software systems enable the collection of usage data at a massive scale. At same time, scale and scope processes have grown substantively. Automated tools are needed to increase speed quality controlling processes. The has great potential as driver for However, research still lacks constructs collecting, refining utilising in objective this paper is introduce framework data-driven engineering. Undulate covers generating, collecting from business supported by produced. In addition, we define concepts process extreme continuous experimentation an exemplar engineering process. We derive requirements literature, with focus on papers inspired practical problems. apply multilevel modelling language describe related experimentation. give provide overview data, augmenting it additional dimensional aggregating along dimensions computing different metrics based other metrics. represents significant steps previous insight towards standardised engineering, enabling application soft methods artificial intelligence. • constructed Undulate, novel systematically processes, storing database, thus creating basis defined general principles how can be used exercise automated control decision support such extended applied key scientifically rigorous conceptual framework.