"Blind audition" of resumes

[erinmfarr]

Theme:
Problem Statement 1 - data driven recruiting
(secondary) Problem Statement - remove bias from Machine Learning models

Brief description of your idea:
Create a tool that scans a resume to automatically mask words that enable recruiters to be susceptible to bias, such as:
• Name
• Gender-related words
• College
• Age
• Address
• Hobbies

This tool can also be run before a resume is inputted into common machine learning algorithms, so such bias isn't introduced into a company's recruitment models.

Performing this action before a resume is seen by a human can help reduce the impact of implicit bias from a recruiter (or at least, get it past that recruiter to a next step.)

Companies using this technology can also advertise they have bias-free recruitment tooling to help ensure the most qualified candidates are hired.

What makes your idea unique?:
A cursory search didn't find an existing tool specific to scanning resumes to strip bias, but there may be one out there.

What would be the impact of your idea if implemented?:
Recruit the best talent by removing bias from the decisions, which then increases opportunities for underrepresented minorities.

Skills to contribute (e.g. development, architecture, research, design or anything else):
architecture, design, possibly development (pending tooling setup/language)

[josiemundi]

This is a cool idea! However, it does already exist to some degree. Blind hiring is already something that a number of companies do already. I found a couple of companies that offer recruitment and applicant tracking software that masks certain information from the recruiter that could lead to bias.

I think you would probably need it to be combined with some sort of tool that could run through the entire process, otherwise the bias could potentially just end up being delayed to a later stage.

You could have some sort of tool that informs the recruiter they may be making a biased decision during the process if they discard a candidate cv that has similar skills to another and ask them to validate their decision. This could help make them more accountable for their decisions. Perhaps prompt for data (e.g data from recruitment tests or from cv).