Skip to content
Open access

EduArchive AI 2.0: An AI-Powered Academic Intelligence and Previous Year Question Paper Analytics Platform

Oct 2026 · International Journal of Creative and Open Research in Engineering and Management · 0 citations

Abstract

Previous Year Question Papers (PYQs) are valuable academic resources that help students understand examination patterns, identify frequently asked questions, recognize important topics, and plan their preparation effectively. However, conventional question-paper repositories mainly provide document storage and downloading facilities, requiring students to manually browse and compare papers from different years. This process is time-consuming and makes it difficult to identify repeated questions and important topics systematically. To address these limitations, EduArchive AI 2.0 is proposed as an artificial intelligence-powered academic intelligence and question-paper analytics platform. The system integrates Optical Character Recognition (OCR), Natural Language Processing (NLP), sentence embeddings, semantic search, Large Language Models (LLMs), and Retrieval-Augmented Generation (RAG) to transform historical examination papers into a searchable academic knowledge base. Its multi-agent architecture consists of five specialized agents: Ingestion, Analysis, Search, Recommendation, and Admin Intelligence. The platform supports processing digital and scanned PDF documents, detecting semantically similar questions, identifying frequently occurring topics, classifying question difficulty, and retrieving relevant academic information through natural-language queries. The RAG-based “Ask Your Papers” assistant is designed to generate responses using relevant content retrieved from uploaded question papers. The proposed implementation uses React.js, Python Flask, Sentence Transformers, ChromaDB, LangChain, Redis, and Docker. EduArchive AI 2.0 aims to reduce manual document analysis, improve access to academic information, and support more organized examination preparation. Its effectiveness should be evaluated using representative question-paper datasets and measurable performance metrics

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.