The contemporary talent acquisition landscape demands scalable, objective, and computationally efficient screening solutions capable of processing large volumes of candidate applications without sacrificing evaluation quality. This paper presents an AI-powered smart recruitment and resumes screening system that automates the initial candidate screening phase through Term Frequency-Inverse Document Frequency (TF-IDF) vectorization and cosine similarity computation. The system is engineered as a full-stack web application using the Django 5.x framework, Python 3.10, and SQLite3, implementing a dual-role user model that clearly segregates recruiter and candidate workflows under role-based access control. Resume content is extracted from candidate-uploaded PDF documents using the pdfplumber library, transformed into TF-IDF vector representations via scikit-learn, and matched against structured job requirements to produce a quantitative percentage match score. The resulting ranked candidate presentation enables recruiters to prioritize review of the most qualified applicants within seconds of application submission. Experimental evaluation demonstrates that the proposed TF-IDF cosine similarity approach achieves a match accuracy of 76.5%, significantly outperforming conventional keyword Boolean filtering (47.1%) and comparable to manual human screening (59.9%) while operating at a fraction of the time cost. The responsive Bootstrap 5 interface supports seamless interactions for both user roles across desktop and mobile environments. The system establishes an open-source, interpretable, and extensible foundation for intelligent recruitment automation, with future pathways incorporating transformer-based semantic matching, named entity recognition, and cloud-native deployment architectures.
Bhanu et al. (Thu,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: