{"id":11253,"date":"2025-05-19T13:06:54","date_gmt":"2025-05-19T13:06:54","guid":{"rendered":"https:\/\/www.gsma.com\/get-involved\/gsma-foundry\/?post_type=gsma_theme_resources&p=11253"},"modified":"2026-10-27T21:43:04","modified_gmt":"2026-10-27T21:43:04","slug":"china-mobile-customer-service-large-model","status":"publish","type":"gsma_theme_resources","link":"https:\/\/www.gsma.com\/get-involved\/gsma-foundry\/gsma_resources\/china-mobile-customer-service-large-model\/","title":{"rendered":"China Mobile’s Customer Service Large Model"},"content":{"rendered":"\n

Challenge\/Problem Being Addressed:<\/strong> In the digital economy era, the widespread adoption of digital technologies in the service sector has resulted in applications like smart customer service and customer experience management systems, which have improved service quality to some extent. However, due to technological limitations, persistent issues such as irrelevant responses, slow service response times, and inefficiencies in problem resolution remain challenging, making it difficult to provide satisfactory customer service.<\/p>\n\n\n

\n
\"A<\/figure>\n<\/div>\n\n\n

Action, Solution, Operation:<\/strong> China Mobile developed the Customer Service Large Model which is built on a standard network interconnection model, leveraging an AI + X systematic AI development approach. It employs a training paradigm of “pre-training + instruction fine-tuning + reinforcement learning with human feedback”. This model also integrates data application resources from customer service information systems. By focusing on data insights, technological capabilities, and industry applications, it promotes the training & inference and application of the Customer Service Large Model through an efficient collaboration between the online marketing service center and the Jiutian Innovation Research Institute, thereby enhancing smart customer service capabilities.<\/p>\n\n\n\n

Impact:<\/strong> The implementation of the Customer Service Large Model in enhancing smart customer service scenarios has been a gradual journey. It has drawn strength from the independently developed robust smart interaction foundation, a rich professional dataset, and continuous iterations of the reinforcement learning technology. Through multiple version iterations and extensive internal testing, the model has continually improved its accuracy, completeness, and friendliness. This progress has driven diverse value in technological innovation, flexible customization, and emotional empowerment, providing customers with thoughtful, humanized, and personalized service experience.<\/p>\n\n\n\n

Topic:<\/strong> Customer experience<\/p>\n\n\n\n

MNO:<\/strong> China Mobile<\/a><\/p>\n\n\n\n

<\/div>\n \n