Functional Architecture Optimization of the Chaoxing Learning Platform for Differentiated Instruction Design under the AI-Enhanced Learning Paradigm

2025 marks a period of an explosive growth for generative artificial intelligence, and further accelerating the informatization of education. In this way, personalized learning is becoming increasingly possible. However, the true implementation of personalized learning relies on the technical support of online learning platforms for differentiated instruction. This study examines the limitations of the Chaoxing Learning Platform, such as uniform content delivery, incomplete learner profiling, manual task scheduling, and standardized evaluation which effectively hinder differentiated instruction. Through system modeling and functional analysis, this paper proposes a five-layer architectural optimization framework including Data Layer, Learner Profile Layer, Recommendation Layer, Task Scheduling Layer, Evaluation Layer in order to enhance the platform's adaptability to diverse learning needs.