-2

Tried to automate my resume screening with a tool and it backfired hard

I used this AI resume parser last week to filter 150 applicants for a junior graphic designer role in Austin. It flagged 40 people as "low match" based on keywords, but when I actually read them, three of those had the exact portfolio experience I was looking for. The tool missed them because they used "Photoshop" instead of "Adobe Creative Suite" in their summary. Has anyone else had a filter tool bury good candidates like that?

3 comments

3 Comments

grantt112mo ago

Actually read an interesting piece in the Harvard Business Review about this exact problem. They found that automated filters miss about 25% of qualified candidates because people use different words for the same skills. I have a buddy who runs a small design studio in Portland and he told me he stopped using filters entirely for creative roles. He said the software kept passing over people with really strong portfolios just because they didn't use the right industry buzzwords. Sounds like you found yourself in the same trap with that Photoshop vs Creative Suite thing. Might be worth doing a quick manual scan next time, at least for the first 50 or so resumes to see what the tool is actually catching.

cameronb522mo agoMost Upvoted

Yeah, my friend at a tech startup had that happen. They were looking for a project manager and the filter was set for "Agile." It skipped a woman who had run projects using "scrum" and "sprints" for years at a big company. Her resume just said she "led fast-paced software teams." They only found her because a recruiter did a second pass by hand. How many good people get lost like that?

hayden_rivera1mo ago

Respectfully disagree here. Filters catch way more bad applicants than they miss good ones, and the ROI of manually scanning every resume is terrible when you get hundreds for one role. Someone who can't match their resume to the job description probably won't adapt well to corporate processes anyway.