RevisePlus

Creating subject-specific assessment questions is a highly time-consuming task for educators. Traditional study materials lack adaptability and often fail to provide students with quick feedback or detailed explanations. Educators spend countless hours manually crafting questions, while students are held back by the limitations of rigid, non-customizable resources that cannot adapt to individual knowledge gaps or learning speeds.

We built RevisePlus - an application that incorporates specialized machine learning models to generate subject-specific questions directly from uploaded study materials. Users can input content in any form, including PDFs, raw text, and images. The platform automatically translates generated questions into shareable Google Forms and quizzes for easy distribution, supports native language processing including Sinhala for broad accessibility, and allows users to export tailored questionnaires as formatted PDF documents. To deliver highly accurate text processing while minimizing the costs of large LLMs, we engineered a specialized backend on Flask and Python, utilizing NLTK for robust text processing and efficient transformer models via Hugging Face - integrating with Google Cloud to automate Google Forms generation.

RevisePlus significantly increases productivity by saving valuable time for both educators and students. Beyond workflow optimization, it offers an eco-friendly and cost-effective alternative to purchasing physical model paper books. The platform is designed to be accessible to anyone, ensuring high-quality assessments are available regardless of an individual's technical expertise or English proficiency.