Summary

Hiring is broken. Automated systems reject people who don’t match a keyword template, even when they are more qualified than the "perfect" candidate on paper.

75% of qualified applicants are rejected before a human ever sees them.
88% of employers admit they lose strong candidates due to ATS limitations.

I built Qualifai during the 2025 UCLA Ethical AI Hackathon to rethink how we surface talent. It uses a custom semantic reasoning pipeline built on natural language understanding, designed to extract actual meaning and stories. Unlike legacy systems, it identifies cross-domain skills, infers context, and decouples applicant value from formatting.

Only 12% of companies use semantic search in hiring, yet those that do see a 62% increase in quality of hire. This is the gap Qualifai is built to close. This approach could easily be worked in to existing legacy ATS systems.

I defined the problem, architected the system, designed the product, and deployed it in Replit. Oli supported with research.

Built in one week to challenge how we use AI in hiring. We all deserve more than simple pattern-matching.