Dillon Chi

Dillon Chi

UX Researcher and AI Generalist

Services → About →
Case Study machine-learningsystemic-bias

Auditing Hiring Algorithms for Fairness

90% of Fortune 500 applicants are judged by a machine (an ATS). A toolkit to start testing those judgments and make hiring fairer.

Cover image for Auditing Hiring Algorithms for Fairness
My role
UX Researcher · Systems Design
Client
Speculative / ArtCenter
Year
2021

01 · Data

90% of Fortune 500 companies will judge you before a human does.

All job seekers pass through an Applicant Tracking System first.

The same qualifications. Different outcomes.
Problem Context
Context
The Problem

Applicant Tracking Systems score and filter resumes before any human reviews them. The criteria these systems use — keyword density, school names, address, formatting — can systematically disadvantage candidates who are equally or more qualified.

Evidence
What we built

700 equally-skilled resumes — varying name, institution, zip code, and formatting — to give job seekers and researchers a controlled corpus for testing Applicant Tracking System behavior.

Takeaway
Why it matters

You can’t audit a black box you can’t see. Making the test set public means anyone can start generating evidence — not just researchers with institutional access.

Check out the kit

The Applicant Tracking System (ATS) Testing Kit is publicly available. Run your own resume against the corpus and start collecting evidence.

Check out the kit →