@random_walker I’m the co-founder and CEO of @pymetrics. Repeated peer-reviewed studies show current hiring practices are discriminatory and ineffective. bit.ly/33TMzri, bit.ly/346kZat. We started pymetrics to address this.
pymetrics took tools from the peer-reviewed lit that measure cognitive, social, and emotional features, and are demonstrably related to job performance. We repeatedly audit our platform for adverse impact before and during its use: open-sourced on GitHub: bit.ly/355jjhg.
We see improved retention/performance and diversity: more women, people of color, and socioeconomically-disadvantaged people being hired. pymetrics has presented fairness and validity results at multiple conferences and is in the process of peer-reviewed submissions.
We’ve been joined by former regulators from the EEOC who were tired of investigating employers for discrimination and wanted to be part of the solution, helping us build assessments that are both fair and valid.
We've worked with academics to secure grant funding to study our system, including engaging in a 3rd party audit. We believe stronger guidance and guardrails are needed to ensure compliance with fairness, accountability, and transparency for AI in hiring.
I'd also be happy to discuss all of this in greater detail with you offline.
