image1
Artificial intelligence has rapidly transformed hiring practices, promising greater efficiency and objectivity. Approximately 75% of Fortune 500 companies now integrate AI tools like GPT-based models into their recruitment processes. However, this seemingly objective technology may unintentionally reinforce existing human biases, disproportionately disadvantaging candidates from historically underrepresented groups.
These tools don't create bias; they reflect and magnify our own.
While AI systems can quickly evaluate résumé elements such as names, educational institutions, and employer brands, they often amplify biases related to race, gender, and socioeconomic background. Everything presented in Beyond the Résumé is based on findings and data from:
Beyond the Résumé invites us to rethink how we approach hiring. In a world where résumés, years of experience, and formal education often serve as the primary measures of a candidate’s worth, Beyond the Résumé challenges us to look beyond these traditional markers and focus on the skills and competencies that truly predict job success. Beyond the Résumé promotes evaluating candidates based on practical skills, better predicting job success and promoting inclusivity.
image
Explore detailed résumés from six qualified candidates.

Hover over highlighted elements to see how details like names, universities, employers, locations, and technology choices shape AI evaluations, influencing candidates from the very first screening.

Suni Tran

Candidate with Asian female identity

123 Main Street, Anytown, USA

Email: suni.tran@example.com

XYZ University — B.S. Computer Science, GPA: 3.8

Experience

Google, Software Engineer (2018–Present)

  • Designed and deployed scalable backend systems
  • Led cross-functional team of 6 engineers
  • Received internal award for innovation in AI safety tooling

Skills

Python, Go, Kubernetes, AWS, GCP

Score: 78

Darius Mosby

Candidate with Black male identity

123 Main Street, Anytown, USA

Email: darius.mosby@example.com

UCLA — B.S. Computer Science, Class of 1998

Experience

ABC Corp, Senior Developer (2018–Present)

  • Refactored monolithic app into microservices
  • Mentored 4 junior developers
  • Championed diversity hiring initiatives

Skills

Java, Spring Boot, SQL, Docker

Score: 70

Isabella Garcia

Candidate with Hispanic female identity

456 Lake Drive, Austin, TX

Email: isabella.garcia@example.com

University of Texas — B.S. Information Science, GPA 3.7

Experience

Meta, Frontend Engineer (2019–Present)

  • Built modular components with React
  • Improved ad dashboard performance by 45%
  • Hosted internal UX diversity workshops

Skills

JavaScript, TypeScript, Figma, Jest

Score: 74

Fatima Habibi

Candidate with Middle Eastern female identity

678 Oak Lane, Chicago, IL

Email: fatima.habibi@example.com

University of Illinois — M.S. Data Science

Experience

Capital One, Data Analyst (2020–Present)

  • Built interactive credit risk dashboards
  • Automated monthly reporting using Python
  • Collaborated with compliance on fairness testing

Skills

Python, R, Tableau, SQL

Score: 72

James Smith

Candidate with White male identity

999 West Blvd, San Francisco, CA

Email: james.smith@example.com

Harvard — B.A. Economics, GPA 3.9

Experience

Stripe, Product Manager (2019–Present)

  • Led launch of ML-driven fraud detection
  • Presented at 3 global fintech conferences
  • Managed team of 12 product and design staff

Skills

SQL, Product Strategy, A/B Testing, JIRA

Score: 90

Amina Yusuf

Candidate with Black Muslim female identity

45 North Cedar St, Dayton, OH

Email: amina.yusuf@example.com

Al-Quds University — B.Sc. Software Engineering

Experience

UNDP, Technology Fellow (2017–2021)

  • Deployed humanitarian logistics platform across 3 countries
  • Led development team for refugee data system
  • Trained local staff in agile methods and security protocols

Skills

Python, PostgreSQL, Docker, Django

Score: 69

Subtle differences in how résumés are evaluated—based on names, schools, employers, or locations—can influence outcomes from the start.

Follow how AI-driven evaluations shape each hiring stage—screening, shortlisting, and final offers—as small biases compound over time, favoring some candidates while sidelining others.
All six candidates submitted résumés. GPT-based hiring models immediately evaluate features like name, degree, graduation year, and employer brand. According to the paper, GPT-4 rated White male résumés significantly higher than others, even when experience and education were equal.
Darius and Amina are deprioritized at the screening stage. Armstrong et al. found that Black male and Muslim-sounding names experienced a 10–20 point reduction in model-generated scores across otherwise equivalent résumés.
The interview shortlist includes James and Isabella. Isabella's inclusion highlights how working at a prestige firm like Meta slightly mitigated identity bias. James consistently benefited from identity, employer, and education signals.
James receives the final offer. The study found White male candidates were 2.75× more likely to receive job offers in LLM-based scoring pipelines, underscoring how compounding privileges result in biased hiring outcomes.
Early biases in hiring don't just affect who gets the job—they set the trajectory for long-term career and salary growth.

Projected earnings over time show how initial advantages or disadvantages widen, compounding gaps with every job move and raise.
Projected compensation reflects systemic disparities embedded in AI scoring models. The initial salary projection assumes $150,000 × (score/100), with a 3% annual raise.
Every three years, candidates are assumed to change jobs. When changing jobs, bias is reapplied, compounding advantages or disadvantages over time.
Amina and Darius, despite strong technical backgrounds, are projected to earn significantly less over 10 years. Bias during initial hiring and subsequent job changes reduces their total earnings substantially.
Fatima and Suni, despite prestigious employers or advanced degrees, still face penalties over time due to race, gender, and region biases—affecting salary growth at every career move.
James Smith benefits the most from positive bias across all stages, leading to the highest cumulative earnings after 10 years.

Formula: Start salary = $150,000 × (score / 100); +3% raise/year; bias reapplied at job changes (every 3 years).

Small hiring biases snowball into major career earnings gaps over time. What starts as a few points' difference can grow into six-figure disparities.

Use the calculator to see how attributes like race, gender, education employer, graduation year, and English fluency shape AI evaluations long before interviews begin.

Bias Calculator

Select candidate attributes to simulate AI score:

Estimated Score: 85

Take a step beyond the résumés. See talent differently.
Find people with real potential who don't fit the usual mold
Build teams that see problems from different angles
Improve results through more complete perspectives
Fill positions faster by focusing on what the job actually needs