The idea behind the Fit-Finder

In summer 2025, a group of two students led by a professor at Sacred Heart University in Fairfield (CT) started working on a research idea. Is it possible to create an AI powered tool that can score the quality of a candidate's resume against a job description, ensuring fair hiring practices by comparing the candidate's skills and qualifications to job requirements instead of just working on documents similarity ?

Over the past decade, companies have adopted Applicant Tracking Systems (ATS) to streamline recruitment. However, traditional methods that are using keyword matching and simple ML models often miss qualified candidates and can perpetuate hiring bias—as seen in in Amazon's gender-biased recruiting engine.

With LLMs becoming mainstream, candidates increasingly use AI to artificially match their resumes to job descriptions. This creates a new challenge, distinguishing genuine qualifications from almost fully AI-generated content designed to bypass ATS systems and new opportunities.

The core of our research is to demonstrate the LLM superiority by showing how LLM-based matching models outperform classic approaches in understanding context and requirements. We investigate how LLM-based algorithms can provide fairer candidate evaluation compared to traditional methods, while delivering transparent insights into why candidates match or don't match roles, supporting better hiring decisions.

For companies, this means reducing hiring bias, detecting fraudulent applications, and making faster, more informed decisions with explainable AI scoring. For candidates, it provides transparent feedback on how qualifications match job requirements, with clear insights into strengths and gaps. For researchers, it advances the field of fair AI and explainability in hiring, demonstrating LLM capabilities and limitations.

We are now in the phase of collecting data to train our models to push the research to the next phase. If you are interested in voluntarily participate to this data collection and score your resume with whatever job description you provide, please login and get started. Thanks!