StudyVault: A Grounded Study Assistant for Citation Backed Learning
Course Instructor
Pramod Gupta
Abstract
StudyVault is a grounded study assistant built to help students learn from their own course materials with greater accuracy, traceability, and focus. The system lets users upload documents such as lecture notes, slides, and readings, then processes them through a retrieval based workflow that chunks, embeds, and indexes the material for evidence driven question answering. Unlike general AI study tools, StudyVault emphasizes citation backed responses, allowing students to inspect the exact source passages used in each answer. In addition to chat, the platform generates flashcards and quizzes from selected documents or topics, tracks quiz attempts, and identifies weak areas through topic level mastery signals. Built with a FastAPI backend, React frontend, SQLite metadata storage, and vector retrieval, StudyVault shows how grounded AI can support trustworthy academic study by combining document understanding, adaptive practice, and source linked review in a single workflow.
StudyVault: A Grounded Study Assistant for Citation Backed Learning
StudyVault is a grounded study assistant built to help students learn from their own course materials with greater accuracy, traceability, and focus. The system lets users upload documents such as lecture notes, slides, and readings, then processes them through a retrieval based workflow that chunks, embeds, and indexes the material for evidence driven question answering. Unlike general AI study tools, StudyVault emphasizes citation backed responses, allowing students to inspect the exact source passages used in each answer. In addition to chat, the platform generates flashcards and quizzes from selected documents or topics, tracks quiz attempts, and identifies weak areas through topic level mastery signals. Built with a FastAPI backend, React frontend, SQLite metadata storage, and vector retrieval, StudyVault shows how grounded AI can support trustworthy academic study by combining document understanding, adaptive practice, and source linked review in a single workflow.
