Authors:
R. Vinoth, Adapa Brunda Mani, Som Satyesh Behera, S. Nahul Rathinam, Azlin Abd Jamil, Melanie Elizabeth Lourens, Zhang Min, Liu Yan

Addresses:
Department of Computer Science and Engineering, Faculty of Engineering and Technology, SRM Institute of Science and Technology, Ramapuram, Chennai, Tamil Nadu, India. Faculty of Management, Universiti Teknologi Malaysia (UTM), Johor Bahru, Johor, Malaysia. Faculty of Management Sciences, Durban University of Technology, Durban, KwaZulu-Natal, South Africa. Department of Sales, Huarong Group, Jiangsu, China. Department of Sales, Pearl River Exports Co., Guangdong, China.

Abstract:

The World’s Employment is in the grip of a paradigm shift with Human-Machine Convergence. In addition, researchers observe that Generative Artificial Intelligence (GenAI) and automated workflows are transforming the nature of technical roles across every major industry vertical; a clear disparity has emerged between traditional academic curricula and the ever-evolving competencies demanded by modern employers. This imbalance creates structural inefficiencies for both job seekers, who lack actionable career cues, and organizations with longer time-to-hire cycles due to incompatible candidate profiles. This coming schism is addressed by the next-generation multi-agent computational framework Job Trend Analyzer and Skill Gap Predictor. The system can continually aggregate real-time job data from LinkedIn, Indeed, Naukri, and technical portals using huge, high-frequency web-crawling agents. Keyword-based techniques cannot identify subtle skill needs, seniority indicators, and wage benchmarks as effectively as Large Language Models (LLMs) do. K-means clustering for professional role grouping to reveal hybrid occupations, Multiple Linear Regression (MLR) for financial income modeling, and a cosine-similarity-based semantic engine for targeted skill-gap analysis comprise the underlying analytic architecture. Named Entity Recognition (NER) pipelines extract technological stack attributes, and a bias mitigation layer checks job descriptions for gender-coded language. An experimental 30-day study of over 150,000 Indian and UAE job ads demonstrated 94.2% automated skills extraction accuracy, an F1-score > 0.95 across major professional disciplines, and a 35% increase in candidate interview success. It cut career trend analysis time from 48–72 hours to 4.2 minutes.

Keywords: Job Market Analysis; Skill Gap Prediction; Large Language Models (LLM); Predictive Analytics; Web Scraping; Career Development; Named Entity Recognition (NER).

Received on: 08/09/2025, Revised on: 01/11/2025, Accepted on: 28/12/2025, Published on: 19/08/2026

DOI: 10.69888/FTSFDS.2026.000739

FMDB Transactions on Sustainable Finance and Data Science, 2026 Vol. 1 No. 3, Pages: 117-130

• Views : 47
• Downloads : 11

Download PDF