I wasn’t clear. I know they can include additional simple factors like MCAT score, race and non-trad status - things that are objective and easily coded. I was specifically referring to the idea of the filter knowing more nuanced EC/circumstantial type things e.g. student worked 30 hours/week to afford college tuition or applicant was primary caregiver for parents throughout college, both of whom died in the middle of the 2nd semester junior year, student grew up in foster care after legally being removed from parent that was sexually abusing them, student founded a now well known patient/disease advocacy group etc. If in fact they can use textual analysis to determine the difficulty of the path the applicant has had (a factor I know we use, but is scored by people) then I guess my school doesn’t have that software. I know our screeners (of which there are many) don’t need to read an application with a subpar GPA/MCAT in much depth (i.e. doesn’t take long) to determine if it’s worthy of moving on to the next step (being read in depth) or not.
I believe therefore, that a lot of middling 3.2/29 white kids (assuming this is within my 3SDs estimate) make it through the auto screen even at top schools and get tossed in the trash after one set of human eyes spends a minute on the app.