Data Scientists


Adam D’Souza

Adam holds a Ph.D. in physics. Besides data analysis, his interests include reading, cooking, playing the violin, and tweeting. He plans to try one or more of these things very soon.

About Centrality

Paul Fairie

Paul completed his Ph.D. in political science at the University of Calgary in 2013, where he has taught local politics and research methods. As @paulisci, he tweets fun charts and bad jokes.

Contact us

Michael Underwood

Michael is a software developer with a doctorate in physics. A Calgary native, he has also lived abroad, and travels extensively for work and play. He enjoys finding solutions wherever problems exist.

Our history and experience

Based in Calgary, Centrality was born of an idea: what if we could harness knowledge from both the social sciences and the natural sciences to help organizations get better value from their data? Adam, Michael, and Paul have been friends for a long time, during which they discovered a shared love for data-driven decision making. Having noticed that many political consultants and think tanks are formed either purely from the ranks of social scientists or purely by programmers, they felt that their diverse skill sets could be effectively combined in the study data.

Centrality’s first project is an ongoing study of the voting patterns of city councillors in municipalities across the country. The data are messy, extracted from city council minutes that are unstructured and riddled with errors, both typographical and substantive. We felt that this study would be an excellent opportunity to demonstrate our skills at scraping, cleaning, analysing and visualising data, and to promote critical thinking about democracy at the same time. CityBlocs has attracted a fair amount of attention in its short lifetime. In addition to coverage from the Hamilton Spectator and from Joey Coleman, an influential urban affairs blogger in Hamilton, we’ve also received coverage from the Torontoist and have been tweeted about by city councillors in Montreal, Toronto, Hamilton, and Calgary (even our own Mayor Nenshi seems to be a fan).

Who is this team?

Paul’s PhD work was in the field of religion and voting in Canada. He wanted to explain the empirically well-known fact that Catholics vote disproportionately Liberal and Protestants disproportionately Conservative, the so-called “Canadian Catholic voter problem”. It turns out that, broadly speaking, it’s because of “frozen cleavages”: religious groups support the political movements that treated them better initially—forever. Before arriving at this conclusion, Paul had to build and test his models using data from every provincial election that has ever occurred in Canada. Unsurprisingly, the results of the 1908 Saskatchewan election (and indeed most provincial elections since Confederation) have not been conveniently digitised and entered into a database; this was a labour of love on Paul’s part.

Adam’s PhD was in quantum information theory, condensed matter theory, and quantum optics. The first half of his graduate studies were focused on measurement-based quantum computing, a field that borrows heavily from graph theory, which also happens to be an extraordinarily useful mathematical framework for the description of social systems. He spent the latter stages of his thesis work generating vast amounts of data from numerical simulations of optical cavities, and learning techniques for summarising and visualising them.

Michael’s graduate work was also in quantum information theory, and its connections to condensed matter and quantum optics. His work made even heavier use of graph theory than Adam’s. Beyond the relatively arcane world of quantum physics, Michael also has extensive experience as a software developer, having worked on a diverse array of projects primarily related to mapping GPS updates from work trucks, commuter-rail trains, and seismic vessels, among others. Along with writing communications software and mapping applications, he has created route-optimisation routines for a delivery service, summary report generating programmes for governmental and private-sector vehicles, and real-time reporting of train ETAs at passenger platforms.