Not a "spell checker" per se, but it suggests word choices that are more welcoming to female and minority job candidates.
So they say. I'd be interested to have that validated (not "proven!") by an independent analysis that is not making money from this. Take JDs before and after making the changes the software suggests, then do a blind test about which version gets more responses. And better responses -- if you are posting a response, you probably want more women candidates without also getting swamped by additional unqualified male candidates. You can imagine that if the main measurement is "how many women apply", you might get a "false positive" by simply tweaking the JD to make the job less demanding. But that's subtly different than getting more qualified female candidates applying.
Add a comment...