In order to better utilize the massive data resources under the smart campus system, a student management system based on artificial intelligence AI and digital technology is proposed. The study, aiming at student behavior analysis, designed distributed data storage management, data mining, and analysis, as well as optimization schemes for data clustering algorithms used in the system development process, and tested the key technical solutions. The experimental results indicate that the error rate of the improved algorithm has decreased by about 4%, and it has higher accuracy in determining data center points. The traditional algorithm has 16, 23, 27, and 36 iterations in four sets of data processing, while the improved algorithm has 15, 17, 20, 22, and 24 iterations in four sets of data processing, respectively. From the results, it is evident that as the data volume grows, the algorithm requires more iterations, leading to longer processing times with larger data sets. Compared to traditional algorithms, improved algorithms have fewer iterations, resulting in faster convergence speed and higher processing efficiency. The effectiveness of the plan has been validated, providing a basic and complete reference for similar applications in other universities.