From Job Postings to Curriculum Decisions: Using AI to Generate Workforce Intelligence for MSW Program Planning

TL;DR

This study demonstrates how AI analysis of job postings can inform MSW program curriculum planning by identifying workforce skills, specializations, and emerging trends, offering a systematic alternative to traditional advisory methods.

Contribution

The paper introduces a novel AI-based methodology for analyzing job posting data to generate workforce intelligence for MSW curriculum development.

Findings

Interpersonal Practice is the most common employment area.

Clinical Assessment and Case Management are key cross-cutting skills.

Trauma-informed care is expanding beyond clinical roles.

Abstract

Social work programs lack systematic methods to align curricula with employer expectations, typically relying on advisory input and alumni surveys rather than direct analysis of workforce requirements. This paper presents a case study demonstrating how one MSW program used artificial intelligence tools to generate organizational intelligence from job posting data for curriculum planning. Using a locally deployed language model, we classified over 40,000 job postings for MSW relevance and alignment with eight practice specializations, then extracted skills, therapeutic modalities, and technology competencies. Interpersonal Practice dominated the employment landscape, followed by Children, Youth, and Families. Clinical Assessment and Case Management emerged as cross-cutting competencies. Macro-level specializations showed co-occurrence patterns among partially aligned positions that largely disappeared among positions requiring MSW credentials specifically. Trauma-informed care appeared in management and evaluation roles, reflecting its expansion from clinical modality to organizational framework. The methodology demonstrates a transferable approach that other programs can adapt for strategic planning, and the findings illustrate the type of intelligence such analysis can yield. The patterns identified entered faculty deliberation as one input among many, interpreted by stakeholders with contextual knowledge no dataset can fully capture.