Implicit bias due to identifying information prevents impartial evaluation of applicants (with both humans and AI)
→ individuals from underrepresented groups are unfairly disadvantaged, despite having the same qualifications as non-minority groups
Bias affects women's access to job opportunities, contributing to:
→ underrepresentation in leadership roles
→ gender pay gap
Introducing Leveler - 🏆WINNER FOR BEST DEI (DIVERSITY, EQUITY AND INCLUSION) HACK - SPONSORED BY FIDELITY @ TechNova 2024
Leveler is a tool that helps remove bias in the hiring process, evaluating candidates solely on their qualifications. Personal information on a resume has been known to introduce gender bias, and we aim to prevent this by providing employers with anonymized resumes that highlight a person's skills without revealing their identity. Our goal is to create a fairer job market where talent is recognized and valued, regardless of demographic factors. By focusing on qualifications alone, we hope to empower diverse candidates and promote a more inclusive workplace.
After signing in, users enter their qualifications and experiences. This data is anonymized to remove information that can reveal a candidates gender, ethnicity, religion and socio-economic background. Their qualifications are matched to the job description and a match score is generated. The anonymized information is forwarded to the employer
Frontend: We developed the client-side of our web application using streamlit, crafting interactive user interfaces and components. Backend: We used PropelAuth for user management and authentication. We used spacy for anonymization and SBERT for resume to job description matching. Text Anonymization: Spacy Resume Matching: SBERT