Canadian companies use machine learning to drive DEI in the hiring process

Knockri, a software company based in Toronto, has developed an AI-powered interview assessment tool. This helps companies reduce bias and enhance Diversity, Equity and Inclusion (DEI) in the hiring process.

Knockri’s interview evaluation tool uses natural language processing (NLP) to evaluate interview transcripts only, overlooking non-verbal cues including facial expressions, body language, or tone of voice. Additionally, race, gender, age, ethnicity, accent, physical appearance, or sexual orientation reportedly did not affect interviewee scores.

Faisal Ahmed, co-founder and CTO of Knockri, says the company takes a holistic and strategic approach to training its models to achieve “objective scoring.” said that , covering a wide range of expressions regarding race, ethnicity, gender, accent, and job duties and choices. After training the model, the company conducts quality checks and adverse effects analysis to analyze scoring patterns and ensure that no quality candidates are overlooked.

While working with high-volume clients such as IBM, Novartis, Deloitte, and Defense Canada, Ahmed says their model can’t analyze every job in the world. “Once we get a new client, new region, new role, or even a new experience level, we wait to push updates, benchmarks, retraining, and scores on that. there is potential.”

To ensure that the data that feeds into the AI ​​isn’t itself biased, Ahmed said he avoids using data from past hiring practices, such as looking at resumes or successful hires from 10 years ago. I am adding. practice. Instead, according to Ahmed, AI models are driven by industrial organization (IO) psychology and focus purely on identifying the types of behaviors and work activities required for a particular job. For example, if a customer service role requires empathy, the model identifies behaviors from a candidate’s past experiences and identifies words that reflect that particular trait, says Ahmed.

He encourages clients to use Knockri at the beginning of the interview process if they have a fairly high volume of applications. This allows us to roll out the same experience, scoring criteria, and opportunities to all candidates.

Ahmed said their technology is meant to help companies lay the groundwork for evaluating candidates fairly and impartially, not to replace human interviewers. Decisions made by Knockri are reviewed by humans, and human interviewers are inevitably involved in the later stages of the interview process.

“We’re not going to solve all your problems, but we’re going to point you in the right direction,” concludes Ahmed.