Note: this is a past event Check out the current event!

Irio Musskopf

Data Scientist at Data Science Brigade


Using Machine Learning and Open Data to Report 216 Brazilian Congresspeople for Corruption
Friday 13:30 - 14:20
machine learning
data science

Your rating:

Tired of seeing Artificial Intelligence applied mostly to advertising but not to benefit the whole population, a group of three Brazilians began in 2016 to analyze open governmental data to detect corruption. The project, named after a Swedish scandal, is called Serenata de Amor Operation. “Serenata de Amor”, Portuguese for “love serenade”, is the name of a popular local chocolate brand. Publishing everything under an open source license the goal is to develop means for any person to detect misuse of public money as small as a candy bar. In a few months of work results account for 216 congresspeople reported for suspicious expenses, few thousand of tax Brazilian Reais returned to the government and more than a million detected for further investigation. With roughly 2k stars on GitHub, many people have been following the project and are anxious to know the next steps. It has been pointed as one of the most prominent Brazilian open source projects in 2016.

As the author of the project, I’m going to walk you through the technology and the context that made Serenata de Amor possible. Since everything is open source, it could be adapted and serve as inspiration for local efforts in the same sense. 

Check the slides


I have been programming for 9 years, 6 of them working in the industry. Passionate about learning, my goal is to solve problems bigger than myself. I believe many solutions can just be found, through experimentation, in hybrids of technology and human sciences.

Lately, you may find me working as the tech leader of Serenata de Amor, a project using artificial intelligence and open data to facilitate social control of the Brazilian public administration. I am also a data science consultant at Data Science Brigade.