EduAdapt AI: Multimodal Adaptive Learning System

Inspiration

Students often struggle to learn effectively because their lecture videos, textbooks, and presentation slides are scattered across different platforms. General AI chatbots may provide answers that are not aligned with their course materials or current understanding. This inspired us to build EduAdapt AI, an intelligent learning companion that makes studying more personalized, reliable, and effective.

What it does

EduAdapt AI combines lecture videos, textbooks, and slides into a unified, source-cited knowledge base. Students can ask questions and receive explanations linked to the original page, slide, or video timestamp. The platform generates adaptive quizzes, evaluates answers, identifies weak topics, tracks topic-wise mastery, and recommends personalized learning activities.

How we built it

We designed a full-stack web application with a modern user interface, a backend API, document-processing pipelines, AI-powered retrieval-augmented generation (RAG), and a database for storing learning materials and student progress. A learner model updates topic mastery based on assessment performance and uses the results to recommend suitable practice questions.

Challenges we faced

The major challenges include extracting information from different file formats, understanding diagrams and figures, preserving accurate source references, preventing unsupported AI answers, generating reliable assessment questions, and updating student mastery consistently.

Accomplishments

Our goal is to bring learning resources together in one platform, provide traceable AI explanations, and make assessments responsive to each student's learning needs. EduAdapt AI aims to help students understand difficult concepts, focus on weak areas, and study more efficiently.

What we learned

This project helped us explore multimodal AI, source-grounded question answering, retrieval systems, adaptive assessment, learner modelling, and full-stack application development.

What's next

We plan to improve diagram understanding, add multilingual tutoring, generate personalized revision materials, introduce spaced repetition, and evaluate the system using benchmark datasets and simulated learner profiles.