As Information Technology continues to develop at a rapid pace, the amount of data generated in various fields is exploding, evaluating data effectively is essential for decision making, quality improvement, and user experience optimization. In this paper, we build a generic evaluation platform for the Web based on deep learning models by combining Python, Flask and deep learning libraries (e.g., TensorFlow or PyTorch). Through this platform, users can upload their own datasets and choose pre-trained models or customized models for prediction and evaluation. The platform architecture is clearly designed to include data processing, model training, evaluation interface and visualization display, The design and implementation process of the platform is demonstrated through actual cases of enterprise customer satisfaction evaluation and teaching evaluation, which provides an efficient and intelligent solution for the evaluation work in the related fields, and looks forward to the future optimization direction.