An AI-based framework for continuous alignment between higher education and industry skill needs

The misalignment between higher education curricula and rapidly evolving industry skill requirements is a persistent concern. This study presents CurriculumCatalyst, a prototype framework that combines natural language processing (NLP) and Retrieval-Augmented Generation (RAG) to support periodic, evidence-grounded alignment of academic programmes with labor-market signals. The framework integrates two textual corpora—a labor-market corpus drawn from public job postings and employment databases, and an academic corpus comprising programme documents and course syllabi—and follows an end-to-end pipeline that (i) extracts structured skill signals, (ii) computes semantic alignment between curriculum and market representations, (iii) quantifies skill-coverage gaps, and (iv) drafts curriculum artifacts (module outlines, learning objectives, reading lists, and assessment prompts) constrained by typical institutional requirements. We report a preliminary, exploratory evaluation conducted on a pilot corpus rather than a confirmatory empirical study; no human subjects, expert reviewers, or baseline benchmarks were included in the present evaluation. In this preliminary run, the prototype surfaced skill-coverage gaps above 20% in selected technology-oriented topics (AI programming, cloud computing, data analytics), which we present as an indicator of the prototype’s diagnostic utility rather than as a generalisable population estimate. The contribution of this work is therefore methodological and architectural: a transparent, reproducible pipeline that translates labor-market evidence into curriculum diagnostics and reviewable curriculum drafts. Pedagogical validity, comparative performance against simpler baselines, and downstream educational effectiveness are explicitly identified as future work and are not claimed by the present study.