Ethical AI

At Knockri, We Build Technology for Good.

We recognize the importance of creating responsible and transparent AI solutions that have a positive impact on people, companies and society overall. Building ethical AI means building our skills assessment tools based on the principles of transparency, careful data selection, bias reduction, verification and validation. As Knockri continues to break new ground and be the front runner in AI assessments, we are steadfast in upholding these major tenants.

We Help Make Better Decisions

Every company has different requirements, and one size certainly does not fit all. That is why Knockri does not just focus on finding you a great candidate, it finds you a great candidate that’s right for you based on evidence.To do this, Knockri merges the practise of Industrial Organizational Psychology and machine learning to shortlist candidates grounded in key performance indicating skills. Skills such as growth mindset, collaboration and empathy that scientifically show strong evidence of good performance within a job.

We provide a high degree of customization where you can quickly derive these KPI skills from your core competency models. Applicants are evaluated on relevant criteria and a defensible scientific framework. An additional advantage of Knockri’s assessment tools is that we assess candidates' skills, not their personality. The AI technology looks at 100% of the incoming applications and rates them against your ideal candidate profile. This makes our solutions more customized to the job. No core competencies or still building? No problem. Knockri will have you covered with our scientifically tested job personas.
Decision making illustration

Removing Algorithmic Bias Is a Top Priority

To ensure our AI is as unbiased as possible, we have created a dataset that is “full spectrum”. Built from scratch, our dataset avoids inherent bias and has high reliability of representation of people. This even includes those with diverse accents and cultures, so our AI is trained based on knowledge that is inclusive and diverse. When quantifying a candidate’s competencies, the algorithm does not account for an individual’s ethnicity, gender, appearance or sexual preference as measures of desirability for a hire.

In comparison to the previous year’s short-list, Knockri surfaced:

13%

Increase in gender diversity

12%

Increase in racial diversity

We Advocate for Equal Job Opportunities

Knockri was founded in 2016 as a solution to our own problems of facing discrimination in the hiring process. The cofounders began their study to create Knockri – a skills assessment tool deeply rooted in I/O Psychology & built from the ground-up with ethical AI and inclusivity in mind. To date, Knockri has assessed hundreds of thousands of candidates worldwide, helping to reduce hiring bias, cut costs to hire and improve candidate experience.

Knockri was awarded as Innovator of the Year by Ascend and has been given a seat at the World Economic Forum’s Global Council on Equality and Inclusion!

We Ethically Collect and Protect Data

We never use open-source data for algorithm training. All data is sourced solely through our platform and gathered through our relationships with enterprise. All data is subject to quality assurance processes to ensure there is no bias due to bad data. Our data is annotated by trained I/O psychologists based on a scientifically validated I/O framework.
At Knockri, we provide a level of security with technical and organizational measures in order to protect our clients and their applicants' personal data.
Data report illustration

We Verify and Validate Regularly

Knockri strictly abides by the Uniform Guidelines, the APA principles, and goes over and above, to support diversity and inclusivity in the design and implementation of AI skills-based assessments. We are experts in the field of employee selection and ensure that our selection procedures are executed reliably and effectively in any organization.

The Knockri assessments are validated using scientific best practices, and are predictive of job performance and other business outcomes. We pride ourselves on the validation of our assessments, and always consider the existing attitudes and commitments of organizational leaders when aligning our validation efforts with strategic objectives.

See What Clients Have to Say About Us

"In the new world of AI, we trust Knockri's ability to combine the right mix of science, theory, and technology to create inclusive assessment experiences with results predictive of success at IBM."

Robert Gibby, PhD

Chief Talent Scientist, Society of Industrial Organizational Psychology Fellow, IBM

A.I….Is It Racist?

A.I. is all the new buzz! It’s the super intelligent technology that is claimed to solve our problems. But can bias be one of them?

Many are worried that A.I technology will take a life of its own and make the wrong decisions. Are we in store for a Robot disaster in the making? Read below to find out what Faisal Ahmed, Knockri’s CTO and Co-founder, has to say about how A.I. actually works and how can we be sure that A.I. is actually unbiased!

Faisal Ahmed:

A lot of people ask this, and it’s a valid question. Can A.I. be biased? The answer is both yes and no. A.I can be biased if the data used to train the A.I. is biased. Basically, your results will only be as good as your data!

To ensure our A.I. is unbiased, we have created a dataset that is “full spectrum”. Built from scratch, our dataset is free of inherent bias and has equal representation of all types of employees. This even includes people with diverse accents and cultures, so our A.I. is trained based on knowledge that is inclusive and diverse and nature.

So when quantifying a candidate’s competencies, the algorithm doesn’t account for an individual’s ethnicity, gender, appearance or sexual preference as measures of desirability for a hire.

With Knockri, you can hire your star candidate based on merit alone and nothing else.