This research aims to cope with the following educational challenges which are investigated by the PPS in data science classrooms. Firstly, learning the complex mathematical and computer concepts in a data science course in English can be challenging for some students for whom English is not native. The inconsistent access to timely academic support diminishes students' self-efficacy due to the decreased engagement in the pace-learning gap. Secondly, due to the limited capacity of the classroom and students’ tight schedules, students have limited opportunities for practical project collaboration. Less experience coordinating projects in and out of the classroom leads to the fact that few classes integrate pre-determined critical thinking skills and problem-solving skills due to students' limited engagement and resulting in the limited critical thinking skills. This project will address pedagogical challenges in a data science course through the implementation of a GPT-based Intelligent Feedback & Tutoring System (GIFTS) for both in-class and after-class activities. The proposed GIFTS framework comprises two key components: (1) In-class dynamic GPT-based curriculum content delivery and real-time feedback driven by sequential prediction models optimizes the in-class study progression of data science concepts; (2) After-class GPT-driven recommendation systems that align with students' learning proficiency levels will facilitate round-the-clock personalized support, addressing the scheduling constraints that typically hinder peer collaboration and instructor assistance.