Hyman: Training hiring models on CVs and incumbents amplifies bias
Barb Hyman · She Hit $6m Bootstrapped for HR Tool, Will Place 200,000 Candidates This Year · May 28, 2022 · at 4:15
Barb Hyman, founder of AI hiring platform Sapia.ai, analyzes the architectural flaws in resume-based predictive hiring algorithms.
0:00 / 0:08exact quote · 8.2s
720p mp4 · rendered on demand · StarZero watermark
“When you're hiring off your incumbents, you risk amplifying existing biases, and secondly, when you're using CV data, you're very likely to amplify existing biases.”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →
More from Barb Hyman
Hyman: Google and IBM Watson lack data to match Sapia.ai
“It's something that IBM tried to do with Watson for a couple of decades, but couldn't because they didn't have the data. And even though Google has 10,000 PhDs working in NLP, they can't do it either because they don't have the data.”
Hyman: Companies without data scientists on LinkedIn are not doing real AI
“And one quick check that I suggest to businesses is go out on LinkedIn and see whether there are any data scientists. If they're not, then there's not really AI going on.”
Hyman: Sapia.ai costs $20 per hire compared to $1,500 for RPOs
“So if you're using an agency, I don't know what it's like in the U S but in Australia, you know, you might be paying 1502 thousand per hire for an RPO. You know, you're paying 20 bucks per hire with our technology.”
Hyman: Sapia.ai uses no open-source algorithms or products
“Everything we do is proprietary. So we don't use any open source algorithms or products.”