THE ECONOMIC EFFICIENCY OF AI-DRIVEN TALENT ACQUISITION

Investigated the critical inefficiency of manual recruitment in the technology sector, where HR professionals spend over 50% of their working time on routine document processing, causing recruiter fatigue and hiring errors costing up to six months of a developer's salary. Developed an automated Decision Support System (DSS) built on FastAPI, PyMuPDF, and the Nemotron-3 large language model, transforming unstructured PDF resumes into a structured Candidate Scorecard with a Match Score, Skill Gap Analysis, and targeted interview questions. Proposed a Speed-Dating Bot module that increases candidate reply rates by 40–60% through NLP-driven personalized outreach. Processing a vacancy with 50 candidates decreases from 25.5 to 2 hours, yielding approximately $700 in savings per vacancy at near-zero operational cost. Confirmed full GDPR compliance through Docker containerization with in-memory-only data processing.