In my thesis, I have developed and evaluated Rational API Designer (RAPID) an open-source assistant that aids in designing non-functional requirements, including security, performance, usability, and others, in the architecture of Web APIs. RAPID has been designed as a conversational assistant and has been implemented and evaluated as a rule-based knowledge-based system in Java. To develop and evaluate RAPID, I have used methods from two fields of Artificial Intelligence and Software Engineering, namely knowledge representation, multi-valued logic, automated inference, and empirical software engineering. Parts of this research are published in the Requirements Engineering journal in two separte papers of WBAPIK: A Body on Sturctured Knowledge on Designing Web APIs and RAPID: A knowledge-based Assistant for Designing Web APIs.

In this research project, I have devised a semi-formal approach that allows to analyze openness requirements in parallel with other critical requirements such as security, privacy and performance in the design of interaction mechanisms with third-party applications and services. The devised approach has been applied and evaluated on the case of designing an open automotive platform in collaboration with a research group at Charlmers University of Technology and Volvo company. Parts of this research received best-paper award at REFSQ 2017 and was invited to the Empirical Software Engineering journal.

In this exploratory research project, I have studied how social and human-related aspects of software ecosystems can be formally represented and modeled, and how this modeling and analysis can be used to improve the design of the socio-technical environment of software development in software ecosystems. The application of the proposed approach has been demonstarted on scenarios from Android and iOS ecosystems.