Every year, U$ 61 billion dollars are diverted to fund corruption in Brazil. If all this money went to a single person, he would rank 8th on Forbes Billionaires List. So, to create awareness about the problem we used A.I and machine learning and created a fictional billionaire owner of the money stolen every year. The A.I. became Ricky Brasil. He was trained to talk about corruption, politics, business and economics and showed Brazilians what corruption really thinks.

To create his appearance, we fed an algorithm with more than 150 photos of corrupt convicts in Brazil, so it could define his appearance based on the average skin tone, eye and hair color, and bone structure. For his personality, we spend 8 months feeding Natural Language Understand and Personality Insights APIs with 500 hours of audio depositions, campaign speeches and news about corrupt politicians, lobbyists and businessman from more than 65 sources.

Then, a biography writer/political journalists along with A.I. specialists worked with Machine Learning and Watson Improvement Studio to train the machine to answer more than 15 thousand questions about his biography, politics, economics, business, law and corruption. All that data became Ricky Brazil. We developed a website where the user could see and listen to Ricky and a mobile app that was used as a microphone to asks questions. This mobile app uses a technology that converts the question audio into text .

The question was sent to the A.I. cloud where Conversation and Natural Language Classifier APIs understands the intent of the question and selects the response from the gathered data. Then, the answer is returned to the person in audio. Ricky was interviewed by journalists, law students, and anti-corruption NGOs representants and had his interviews posted on Forbes Facebook Page and on a website. The campaign had more than 180 millions media impressions and ha 97% of positive response. On 2018 - the year of elections in Brazil, Forbes is becoming a protagonist on the discussion about corruption in Brazil.