AI-Driven Talent Analytics for Strategic HR Decision-Making

Thesis Abstract

The study addresses the persistent challenge of aligning human resource capabilities with strategic organizational objectives in dynamic markets, by leveraging AI-driven talent analytics to inform decision-making processes across recruitment, development, and retention. The aim is to develop and validate an integrated analytics framework that translates workforce data into actionable insights for strategic HR decisions. Specific objectives include (1) to identify key predictive indicators of high performance and retention across job families using machine learning models; (2) to assess the impact of automated talent segmentation on recruitment efficiency and candidate quality; (3) to examine how AI-generated insights influence HR decision-makers’ strategic choices and perceived organizational effectiveness; (4) to evaluate ethical, governance, and bias considerations in AI-enabled HR analytics; and (5) to propose a practitioner-oriented dashboard for real-time strategic HR monitoring. The research adopts a mixed-methods design combining quantitative predictive modeling with qualitative validation and interpretation. The population comprises 12 multinational corporations within the technology and professional services sectors, with a target sample of 1,200 payroll records and performance evaluations spanning three years (2019–2021) and 60 in-depth interviews with HR leaders, talent analytics practitioners, and line managers. Data collection instruments include a structured HR analytics dataset featuring variables on recruitment sources, time-to-fill, training engagement, performance ratings, turnover intentions, compensation, and demographic attributes, alongside semi-structured interview guides and a governance checklist. Quantitative analysis employs regression analysis to identify predictors of performance and retention, survival analysis to model turnover risk, and random forest and gradient boosting methods to derive interpretable feature importances for talent segmentation. ANOVA and multivariate analysis assess differences across departments and job levels, while cross-validation ensures model robustness. Qualitative data are analyzed using thematic analysis to extract patterns related to decision-making practices, perceived usefulness, and ethical considerations, with triangulation performed against quantitative findings. The study also applies the Resource-Based View and Dynamic Capabilities theory to frame how AI-driven talent analytics create unique organizational capabilities and adaptability, and integrates a fairness and accountability framework to mitigate bias in model outputs. Expected findings include (i) identification of a concise set of leading indicators (e.g., skill coverage gaps, time-to-productivity, learning velocity, and predictive turnover risk) that consistently forecast high performance and retention; (ii) evidence that AI-driven talent segmentation improves recruitment quality and reduces time-to-fill by 15–25% without increasing adverse impact; (iii) demonstration that decision-makers using AI-generated dashboards exhibit more rapid and aligned strategic HR decisions, with measurable improvements in forecast accuracy for workforce planning; and (iv) recognition of critical ethical and governance considerations, including data privacy, model explainability, and bias mitigation requirements. The study contributes to knowledge by advancing a theoretically grounded, practically implementable framework for AI-enabled talent analytics that links predictive insights to strategic HR actions, clarifies the role of AI in decision-making under uncertainty, and develops governance mechanisms to ensure fairness and accountability. The concluding section offers evidence-based recommendations for organizations seeking to scale AI-driven talent analytics, including governance structures, data quality standards, stakeholder training, and design principles for user-centric dashboards, as well as directions for future research such as longitudinal impact assessment and sector-specific adaptation. The research anticipates that integrating robust predictive models with interpretive qualitative validation will yield a balanced, reliable approach to strategic HR decision-making, capable of enhancing organizational performance while maintaining ethical integrity.