Digital intelligence to...

From left to right: Amanda Arciero - VP of Operations & co-founder of Airudi, Nancy Laramée - IVADO Director of Partnerships, Gilles Savard - IVADO CEO, Pape Wade - Airudi CEO and co-founder, Yoshua Bengio - IVADO Scientific Director

31 Jan 2020

Automating human resources, but leaving the “human” in

Airudi, founded by two human resources experts, is leveraging machine learning algorithms to reinvent and optimize performance in this area. To build their AI-powered platform, the members of the startup are using thousands of résumés and job postings, along with interview transcriptions, recruiters’ notes,  and results of hiring processes.

Professor Bram Adams, the leader of this research project, trains the algorithm using these data, which have been anonymized. Multiple factors need to be taken into account, not least of which the specific characteristics of each employment category.

“You also need to detect all possible biases, because behind all these data are people,” explains Professor Adams, who oversees the objectivity of the AI tool to eliminate all risks of discrimination.

The tool will appear in the form of a CGI avatar named “Pam,” who will courteously question candidates, also adapting to their attitude: if the person is nervous, the tool will detect this and switch to language designed to put them more at ease. In addition, the algorithm will be programmed to detect which language the interviewee is using and be able to switch languages mid-interview.

“Montréal is a bilingual city, where job postings and résumés are in French as well as English,” Professor Adams notes. “That’s quite rare, and it poses a challenge for our research team. And it’s quite an interesting combination in international terms.”

“It’s interesting because there’s a lot of jargon; a lot of specific and technical terminology. That’s where the learning capabilities of the artificial intelligence tool come in. It has to truly understand the needs in each field to be able to offer the best recommendations.”

 ̶  Bram Adams, Associate Professor
Department of Computer Science and Software Engineering, Polytechnique Montréal

Recruiters needn’t worry, though: Pam isn’t about to put them out of work. Rather, artificial intelligence will up their chances of finding the best candidates, faster.

“Staffing processes today can last up to 56 days! Our platform will speed the process and ensure a much more pleasant experience for candidates,” says Amanda Arciero, Airudi’s co-founder and Vice-President of Operations. The AI solution will perform a preselection and then provide the employer with a detailed list of the profiles offering the best match with the posted position. “There will always be a human making the final decision,” Professor Adams adds.

Airudi is also developing a second component that will promote retention, turning Pam into a specialized assistant capable of answering any human resources–related question employees might have. This project, headed by Professor Foutse Khomh of Polytechnique Montréal, who holds one of the FRQ-IVADO research chairs), will compile all of the company’s business rules along with collective agreements, municipal bylaws and other legislation, to provide a comprehensive decision-support tool for managers, assisting them in matters of salary increases, health and safety management, disciplinary procedures, etc.

By rapidly and efficiently processing many frequently asked questions, Pam will deliver invaluable time savings. “This will allow managers quality time with employees to talk about their professional development, a vital aspect, but one that is rarely covered in companies for lack of time,” Ms. Arciero notes. This feature will likely be of interest to larger corporations (whose HR departments are often overloaded) and SMEs alike.

Thanks to Airudi, artificial intelligence will soon be of service to everyone.

Bram Adams
Associate Professor
Polytechnique Mtl