ExamMind AI: A Citation-Grounded, Retrieval-Augmented Study Platform for Exam Preparation
Year: 2026
Type: preprint
Abstract: ExamMind AI is a citation-grounded, retrieval-augmented study platform built to help students prepare for university and competitive exams using their own notes, textbooks, reference PDFs, and past question papers. Instead of relying only on general AI answers that can miss sources or make unsupported claims, the platform links responses to source details such as document, chunk, page, claim, and supporting snippet. It brings together study chat, smart notes, MCQs, flashcards, timed mock tests, mark-based written-answer practice, topic mapping, formula and definition banks, and performance analytics. This paper outlines the Next.js, React, and TypeScript client structure, Supabase-based data schemas, considerations for the separate RAG backend, and an evaluation framework covering retrieval, citations, faithfulness, answer relevance, abstention, latency, and usability.
Language: English
Subfield: Education
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7237340897
Cites:
Cited by:
Related to:
