Abstract: The railway customer service center has gradually transformed from an initial contact center into a problem solving center and data center, while the supervision and control of massive call recordings still rely on the traditional manual quality inspection mode. Aiming at the problems that exist in the manual quality inspection mode, such as low quality inspection coverage, random sampling result bias, and human subjective factors, which affect the quality inspection results, this paper designed an intelligent quality inspection system for railway customer service based on technologies such as voice recognition, text mining, and big data, to achieve automatic quality inspection analysis of full recording in customer service centers. By providing functions such as intelligent quality inspection, spot check, intelligent learning, and data analysis, the paper compensated for the shortcomings of traditional quality inspection models, and implemented the win-win goal of significantly improving quality inspection efficiency and service quality.