Which candidate is the most “presidential”?

Powered by IBM Cloud and IBM Watson, a Darwin Ecosystem platform analyzes debate performance to gauge personality traits


“We’re taking every single speech by every single presidential candidate, processing it through Watson, and doing personality analysis after every debate,” says Thierry Hubert, Founder and Chief Executive Officer of Darwin Ecosystem.

Analyzing how candidates rank against each other in terms of traits such as openness, assertiveness and conscientiousness, Darwin’s Projected Personality Interpreter (PPI) platform is helping to shape the current political conversation.

“We know some political entities are using PPI to fine-tune how they say things,” says Hubert. “And bloggers reference our analytics to build their arguments.”

Business and academic users can also improve the effectiveness of their communications with the PPI application.

“Whether PPI is analyzing essays, executive speeches or letters to shareholders, it can help answer key questions: Am I projecting success? Am I projecting confidence?” says Hubert.

The glut of data available on any given subject sparked Hubert’s interest in cognitive analytics. “We started by trying to make sense of information overload. How can we understand and use all this data?” says Hubert.

Darwin developed a proprietary algorithm for big data analytics, but it soon made a crucial shift.

“When IBM Watson started to emerge, we made a very serious strategic decision,” says Hubert. “We moved away from our algorithm and Amazon technology and went to the IBM ecosystem.”

Impressed by IBM Watson’s cognitive capabilities on an IBM Cloud platform, Hubert committed to a large-scale transformation.

“We took a year to transform our business and create a platform that would support cognitive solutions at many different levels of complexity,” he says.

Now, the Darwin PPI solution is providing unique insight into the 2016 US presidential race and guiding several Massachusetts police departments in their recruitment processes.

The technology is also helpful in building effective corporate and academic teams.

“There are different types of projects, so how do you know what kind of team is the right team for a given project?” says Hubert. “With our platform, we can determine which team member is the best fit for each role — communicator, project leader and so on.”

For Hubert, shedding light into formerly murky areas is a fascinating line of work.

“In areas like politics, it’s not like reading tea leaves anymore,” he says. “People gravitate toward these solutions because they deliver empirical insight.”