PolyGloT, a novel framework for educators and students, is designed to create, curate, and engage in personalized learning paths in digital education. This open, adaptable, and versatile framework facilitates the design and consumption of adaptive learning experiences. It accommodates diverse educational objectives, encompassing interactive activities and comprehensive assessment tasks. The core of the framework explores the use of generative AI techniques for curriculum creation and exercise curation on the teacher side, and to enhance personalization and feedback on the student side, allowing for adaptations to the varied needs and preferences of individual learners. The research outlines a forward-thinking agenda in AI-driven technologies for education, proposing extensions, envisioning future advancements, and highlighting the challenges in implementing such frameworks. This exploration and demonstration of PolyGloT underscore its potential to advance digital education, aspiring to inspire the education community and stimulate further collaborative research in the development of personalized, learner-centric, and scalable educational methodologies.