Published Online:October 2026
Product Name:The IUP Journal of Computer Sciences
Product Type:Article
Product Code:IJCS041026
DOI:Volume 20, Issue 4, October-December 2026
Author Name:P Rama Santosh Naidu and N V K A Sri Varshini
Availability:YES
Subject/Domain:Engineering
Download Format:PDF
Pages:60-77
Limited Internet access restricts effective learning for students in rural and low-connectivity regions. Most e-learning systems require constant Internet access, offering little support for personalized doubt resolution. This paper introduces ShikshaAI, a fully offline, multilingual AI-powered textbook tutor that transforms standard textbooks into an interactive learning experience. The system uses a local retrieval-augmented generation (RAG) pipeline with vector embeddings and quantized language models to generate accurate, curriculum-grounded responses without Internet dependency. Students can ask questions via text or voice and receive explanations in English or Telugu, with support tools including quizzes, summaries, flashcards, and oral tests. All processing occurs locally, ensuring data privacy and minimal cost. Experimental results on consumer-grade hardware show reliable response times, high contextual accuracy and stable offline performance.
Access to quality education remains highly unequal, especially for students in rural and low-connectivity regions (UNESCO, 2023; World Bank, 2022).