Professor of Computer Science with focus on software engineering
Netflix and Spotify are examples of major software applications. The aim of Benoit Baudry's research is to find ways to build major applications of this type that are resistant to errors and attacks. To achieve this, new mathematical algorithms are designed and experimented with, to increase the systems' diversity and randomness.
Today's systems comprise millions of clones of different software components. Using clones certainly reduces the costs of building large systems, but also increases the risk that millions of copies of the same errors may be further disseminated. Baudry wishes to curtail this risk, while retaining the systems' capacity to deliver service on a global scale.
The research is driven by empirical observations of software and how its quality can be improved. The work of developing, testing and analysing software also takes place in collaboration with researchers at other universities, as well as the business community, within the framework of European cooperation. One example is the STAMP (Software Testing Amplification) project, which aims to develop new solutions for automatic testing of software.