Research Interests

My research interests include model-driven engineering, distributed systems architecture (service-oriented, cloud), software composition, software product lines, software generation and self-adaptive systems. I am particularly interested in software composition paradigms applied to these research domains.

A complete list of my publications is available here: Publications. You can also consult the HAL server, the GLC bibliography tool the Google Scholar search engine or the DBLP index.

Since 2012: Associate Professor (Université Nice-Sophia Antipolis, I3S Lab, UMR CNRS 7271)

2011-2012: Research Scientist (SINTEF IKT)

As a SINTEF research scientist, I was involved into several collaborative research projects: REMICS, ENVISION. In this context, I'm exploring how domain-specific languages and model-driven engineering can tame the complexity of large software design. These research are applied to distributed systems, especially cloud-based ones.

2010-2011: Postdoctoral position (INRIA Lille - Nord Europe)

My postdoctoral studies are focusing on software adaptation and context–aware software product lines. I am working under the supervision of Pr Laurence Duchien (INRIA) and contribute to the Macchiato research project (successor of the awarded Cappucino). This project deals with context–aware computing and software product lines applied to distributed systems (e.g., service– oriented architectures, component based software engineering). The key idea is to consider product derivation paradigm to synthesize the complete implementation of context–aware systems (i.e., structure and behavior), at design time. This method is homogeneously used at runtime to continuously “adapt” the software to its changing environment.

2007-2010: Phd Thesis (Nice-Sophia Antipolis' University, I3S Lab)

My PhD thesis was supervised by Pr Mireille Blay–Fornarino and Pr Michel Riveill, in the context of distributed systems, focusing on Service–oriented architecture and business processes. My main contributions are (i) the definition of a meta–model supporting the design of business process artifacts and (ii) the definition of four order–independent composition algorithms used to support the adaptation of complex business processes, where other approaches (e.g., aspects, features) focus on sequential compositions. A strong execution semantics (based on many–sorted first order logic) is associated to the meta–model, supporting the definition of interference detection mechanisms (e.g., concurrent accesses to a variable). The approach and the associated algorithms are implemented in an open–source tool distributed to four research teams. They were validated on two large industrial case studies. One of this case study focus on the jSeduite system, a distributed open–source information system (based on Web 2.0 principles) developed by the Rainbow research group since 2005 (I’m leading its development team since 2007).

See the Behavioral Compositions in Service Oriented Architecture dedicated page for more information (e.g., abstract, dissertation).