International Journal of Scientific Research and Engineering Development

Graduate unemployment persists globally not primarily because jobs are absent, but because the skills graduates hold and the skills industry requires drift apart faster than curricula and career-guidance systems can track. Existing platforms — job boards, resume-matching engines, and single-model recommender systems — operate on a static, transactional view of the candidate: a resume is parsed once, matched against open postings, and discarded. They do not model how a student's competencies evolve, nor do they close the loop between diagnosis and preparation. This paper introduces SkillSync AI, a multi-agent framework built around a novel construct — the Digital Employability Twin — a continuously updated computational representation of a student's skills, projects, certifications, interview performance, and behavioral signals, maintained across the full duration of their academic and pre-placement journey. Eighteen specialized agents, coordinated by an orchestrator over a shared memory layer, a knowledge graph, and a retrieval-augmented generation (RAG) engine, jointly perform resume intelligence, skill extraction, industry demand tracking, skill-gap computation, learning-roadmap synthesis, interview coaching, and placement-probability estimation. Unlike prior systems that output a ranked job list, SkillSync AI outputs a longitudinal preparation plan that adapts as the student acts on it. We formalize six quantitative constructs — Employability Score, Skill Match Score, Career Readiness Index, Learning Priority Score, Placement Probability, and Industry Relevance Score — and describe the algorithms that update them. A simulation-based evaluation over a synthetic cohort of 1,200 student profiles and 400 job postings, modeled on public labor-market skill taxonomies, indicates that the twin-based roadmap converges skill-gap closure substantially faster than a static-matching baseline, though these results describe a controlled simulation rather than a field deployment. We position SkillSync AI as a research architecture and evaluation protocol for continuous, agentic career preparation, and outline the field-trial work required to validate it with real student cohorts.

Jahnavi Somaraju, Chethan Kumar Reddy P, "SkillSync AI: A Multi-Agent Artificial Intelligence Framework for Dynamic Skill Gap Analysis, Digital Employability Twin, and Personalized Career Roadmaps to Reduce Graduate Unemployment" International Journal of Scientific Research and Engineering Development, V9(4): Page(437-451) May-June 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.