Before joining IBM Research, Michel worked as an IT consultant for three years, mainly in the banking industry. He is currently a Research Staff Member in the Cognitive Systems group at the Dept. of Cognitive Computing & Industry Solutions at IBM Research Zurich, where the driving goal is to help clients gain insight and value from their complex data.
Michel's current work revolves around probabilistic and machine learning approaches to record linkage and entity resolution. A value of an information system compounds with the number of data sources included, but figuring out which pieces of data belong together across data sources is a big challenge; record linkage is a cornerstone in creating these valuable connections.
Previously, Michel studied large volumes of web traffic. His goal was understand browsing dynamics, in order to produce clear pictures of the navigational patterns in a website, and ultimately to provide actionable recommendations for website optimization and marketing purposes.
Other research projects involve business resource allocation, risk management and financial analysis -- most of which he tackles with a combination of probability&statistics, graph theory, optimization and helpful feedback from business partners.