The department where I work in China Mobile mainly provides users with communication services and products. I am mainly responsible for the data analysis and product construction of two product intelligent answering robots and intelligent strategy center.In the process of using the product, users will encounter a variety of call problems, consumption problems, etc., and need a large number of manual customer service to solve the problem. In order to save human resources, intelligent response platform came into being. The intelligent response platform utilizes emerging technologies such as natural language processing capabilities, big data, knowledge graph and artificial intelligence to solve customer problems, deeply integrate business scenarios, accurately identify and solve customer problems, and then provide corresponding whole-process self-service solutions and marketing transformation strategies. While ensuring user satisfaction, marketing is also carried out to ensure the benign operation of enterprises. At present, the average monthly call volume of intelligent response service in China Mobile's online customer service has reached 7000w+, the resolution rate of self-service channels is about 50%, and the average monthly self-service business is about 30w+.However, due to the limited traffic resources, resource allocation and efficient utilization become the focus of the current business, and the policy center carries the traffic allocation work, makes intelligent decisions on user calls, and uses the rule + model capability to achieve real-time and accurate matching between customers and services/resources. Traffic is diverted when the customer service is idle, users with marketing value are diverted to manual precision marketing, traffic is diverted when it is busy, user intentions are simply identified in the rules, and user problems are solved as far as possible in the independent channel in order to reduce traffic pressure.