AI-Powered Web Platform for Automated Curricu lum Vitae Evaluation and Job Matching

Abstract

Traditional recruitment processes rely on manual CV screening, which is time-consuming, inconsistent, and prone to overlooking qualified candidates. This thesis presents the design and implementation of an AI-powered web platform that automates curriculum vitae evaluation and job matching. The system extracts candidate skills from uploaded resumes using named entity recognition, then computes a semantic similarity score between each CV and a target job descrip tion using transformer-based sentence embeddings and cosine similarity. Recruiters are presented with a ranked list of applicants, enabling faster and more objective screening decisions. The plat form is built around three independent components: a Node.js/Express back-end server, a React front-end application, and a Python/Flask AI inference engine. The implemented system was suc cessfully deployed as a live web service and validated through interface testing across all three user roles: candidate, recruiter, and administrator.