Adoption of AI-Based HR Analytics and Its Impact on Firm Productivity, Employment Structure and Wage Dispersion: Evidence from Workforce Data

Artificial Intelligence, Talent Analytics, Employee Performance Prediction, Machine Learning, Workforce Analytics, Human Resource Analytics, Open-Source Datasets, Predictive HR Analytics

Nowadays, artificial intelligence reshapes how HR handles workforce data. This research compares several publicly available workforce datasets to explore whether AI, powered tools predict job performance more accurately. Instead of relying solely on classic statistics, newer machine learning approaches are tested here. Their capacity to outperform older techniques becomes a central point of examination. Evidence, based choices in management gain support when predictions improve. Results hinge on how well these modern models adapt to real, world employment patterns. Starting with raw inputs, the study follows a structured process involving cleaning data, creating features, then applying models to public workforce records containing details on employees backgrounds, roles, involvement levels, and results. Moving beyond basic statistical methods, comparison includes modern approaches, Random Forest, Gradient Boosting, Support Vector Machines, and deep, learning, based neural nets. To judge how well each performs, measures including correctness rate, exactness, completeness, F1 value, along with AUC, guide assessment across trials. What stands out is how AI, driven methods handle prediction tasks much better than older statistical tools, particularly because they capture subtle patterns that traditional approaches miss. Notably strong results come from ensemble and deep learning systems, which maintain consistent precision even when applied to different company environments. It turns out that factors like how involved someone feels at work, how quickly they adapt to new skills, how long they have held their current position, and whether their workload feels manageable play a central part in shaping outcomes. These insights emerge clearly when examining what each variable contributes within the model structure. Despite real, world challenges, the proposed AI, powered talent analytics framework functions as a scalable, data, focused tool companies might apply to track performance, shape employee growth strategies, or spot emerging high performers and those facing difficulties. Insights from this research could assist HR professionals, planners, and executives when embedding intelligent decision aids within workforce design workflows. This work stands out because it draws from several datasets at once, while centering on freely available labor market information, to support results that others can test and extend. Starting where lab, style AI studies often stop, it moves into real HR settings, delivering grounded insights for the growing field of smart hiring systems.