MULTI-STAGE TECHNICAL SELECTION SYSTEM WITH NER-DRIVEN RESUME FLAGGING AND DLIB CANDIDATE VERIFICATION
Publicada el 2026-07-28
Descripción de la oferta
Developed an AI-powered recruitment system using Django and Python to automate the hiring process from resume screening to candidate selection. Implemented NLP and Named Entity Recognition (NER) to extract candidate information from resumes. Applied TF-IDF, Word2Vec, and cosine and to match resumes with job descriptions and rank candidates. Integrated AI-generated resume detection using TF-IDF analysis and GPT-2 perplexity scoring to identify suspicious or AI-written resumes. Built a Random Forest classifier to predict job categories and filter mismatched applications. Developed a secure online assessment platform with role-specific MCQs, coding, aptitude, and subjective questions. Implemented Dlib-based face verification, blink the Evaluated coding challenges using Docker-based sandbox execution and subjective responses using Google FLAN-T5 with semantic similarity scoring. Designed a weighted scoring and candidate ranking system to generate final shortlists based on job-specific evaluation criteria. Built an HR/Admin dashboard.
Skills
Fuente original: freelancer