Presenting a Model for Recruitment Based on a Meritocracy Approach Using Artificial Intelligence (Case Study: Saderat Bank Managers and Employees)

Abstract

The main objective of this study is to design and present a recruitment model based on a meritocracy approach using artificial intelligence to enhance recruitment decision-making in Saderat Bank. This applied research employed a descriptive-analytical design. The statistical population consisted of over 4,000 employees of Saderat Bank in Tehran, from which a sample of 350 individuals was determined using Cochran’s formula. Data were collected from archival sources including recruitment test records, performance evaluations, and promotion scores, along with literature review data. Analytical modeling employed four artificial intelligence algorithms—Deep Neural Network (DNN), Ridge Regression, XGBoost, and Support Vector Machine with RBF kernel (SVM-RBF). Additionally, a Long Short-Term Memory (LSTM) recurrent neural network was applied to detect temporal dependencies and hidden behavioral patterns. Results revealed that across all models, four common predictors—organizational commitment, communication skills, customer orientation, and teamwork—were identified as the most influential components in merit-based recruitment. Performance evaluation scores, professional certifications, and job promotion indices followed in significance. Conversely, ethnicity and political orientation exhibited the least predictive power. The consistency across AI estimators confirms the reliability of the identified merit predictors. Implementing artificial intelligence in recruitment processes significantly enhances transparency, fairness, and precision in personnel selection. By emphasizing behavioral and performance-based criteria rather than demographic factors, AI-driven systems facilitate the realization of genuine meritocracy. Integrating LSTM networks and machine learning frameworks can further optimize large-scale human resource management decisions.