With the economy and society development, big data storage in management has become a problem which can’t be ignored. Task allocation management and optimization is a significant factor in the enterprise sustainable development. The enterprise information intelligent management has become management mode and presents information to the business managers in a more effective and lower cost. In order to solve the management problem under big data and cloud computing, a research on the optimization of intelligent network information management system is proposed. Firstly, build the overall model of big data diversion system under cloud computing platform, and design the architecture of Tiny OS operating system for big data diversion; Secondly, S3C2440 is used as the system control core to design the hardware structure of the shunting system; Finally, the software of the system is developed by using the dynamic loading method of functional subroutines in Linux operating system. The power supply voltage is DC 3.3 V and 1.25 v. DC 5 V is used as the overall power supply of the circuit board, and 0.1 is added at both ends of lm1117 chip F and 100 F capacitance for FIR filtering of power supply. Simulation experiments are carried out to verify the performance, which shows the superior performance of the big data shunting system designed in this paper.