Automated Resume Screening Tool

Abstract

Efficient job–candidate matching is a crucial component of modern recruitment, requiring precision and adaptability to the dynamic job market. This paper presents a comprehensive AI-powered resume screening tool that integrates a resume viewer, an Applicant Tracking System (ATS) score checker, and Zero-Shot Recommendation AI Models. By utilizing state-of-the-art pretrained NLP models such as all-MiniLM-L6-v2, the system processes job descriptions and resumes, transforming them into embeddings and employing similarity metrics like cosine similarity and dot product to compute relevance scores. The ATS score checker evaluates resumes based on job-specific criteria, offering actionable insights for candidates. Testing results show high accuracy in predicting job relevance, with a Top-1 accuracy of 3.35%, Top-100 accuracy of 55.45%, and Top-500 accuracy of 81.11%. These findings demonstrate the tool’s precision, scalability, and efficiency, making it a valuable asset in the hiring process.