StudyBridge AI

A privacy-first AI tutoring app that helps students learn from their study materials through source-linked answers, adaptive quizzes, personalized progress tracking, and targeted recommendations.

Comment

Inspiration

Students often study from scattered notes, textbooks, and lecture materials without having a reliable way to turn those resources into personalized learning experiences. Generic AI tutors may provide answers without sufficient grounding in the student's materials, while traditional quizzes and progress trackers often fail to adapt to individual learning needs.

StudyBridge AI was created to bridge that gap: a privacy-first learning assistant that helps students understand their own study materials, practise important concepts, and identify areas that need more attention.

What It Does

StudyBridge AI transforms uploaded learning materials into a personalized study experience.

  1. Learning from Uploaded Materials

Students can upload supported text files, Markdown notes, text-extractable PDFs, and PowerPoint presentations. PowerPoint ingestion extracts text slide by slide, helping learners work with lecture slides without requiring an additional PowerPoint-processing dependency.

  1. Source-Grounded AI Tutoring

The tutor retrieves relevant passages from the available learning materials and uses source information to support its answers. When the materials do not provide sufficient evidence for a question, the system is designed to avoid presenting unsupported answers as established facts.

  1. Dynamic Practice Questions

StudyBridge AI can generate multiple-choice practice questions from uploaded learning materials, helping students test their understanding of the content they are actually studying. Questions are designed to remain connected to their source material, supporting more transparent revision.

  1. Adaptive Learning and Spaced Repetition

The assessment workflow prioritizes previously missed questions and supports review scheduling. This helps learners revisit difficult concepts instead of spending all their study time on topics they already understand.

  1. Personalized Progress Tracking

The learner model tracks topic mastery and retains historical progress, making it possible to observe changes across learning sessions. The progress dashboard includes SVG sparklines that visualize mastery trends, helping learners identify topics that may need further attention.

  1. Accessible, Mobile-Friendly Learning

A text-to-speech control using the browser's Web Speech API allows learners to listen to tutor responses on supported browsers and devices. A dark/light theme toggle with local preference persistence makes the interface more adaptable to individual preferences.

  1. Privacy-Conscious Architecture

StudyBridge AI is designed around a lightweight application architecture and local development workflow. It aims to keep learning interactions straightforward while reducing unnecessary infrastructure requirements.

How We Built It

StudyBridge AI uses Python and Flask for the application layer, with HTML, CSS, and JavaScript for the browser interface.

Core components include:

  • Document ingestion for supported text, Markdown, PDF, and PowerPoint formats.
  • Retrieval and source metadata for material-grounded tutoring.
  • Assessment logic for practice questions and review prioritization.
  • A learner model for topic mastery and progress history.
  • A progress dashboard with SVG visualizations.
  • Browser-based speech synthesis and persistent theme preferences.
  • Automated tests covering key application functionality.

The project was developed and tested using a mobile-first workflow, demonstrating how a practical educational AI prototype can be built with limited computing resources.

Challenges We Faced

One of the central challenges was connecting tutoring responses to relevant learning materials rather than relying on generic answers. Another was combining document ingestion, assessment, learner modelling, and progress visualization in a lightweight application.

We also needed to make the learning experience useful on mobile devices, where screen space, network reliability, and computing resources can be limited.

Testing and incremental improvements helped us strengthen the application while keeping its architecture manageable.

What Makes StudyBridge AI Different?

StudyBridge AI brings several learning functions together in one workflow:

  • Learn from your own study materials.
  • Ask questions and receive source-supported responses.
  • Generate practice questions from uploaded documents.
  • Revisit missed questions through adaptive review.
  • Track mastery trends across learning sessions.
  • Listen to tutor responses on supported browsers.
  • Personalize the interface with dark and light themes.

Rather than treating tutoring, assessment, and progress tracking as disconnected features, StudyBridge AI connects them into a more personalized study experience.

Testing and Validation

The current automated test run completed successfully:

69 tests passed in 4.91 seconds.

This verifies that the tested application behaviours pass the current test suite. It does not, by itself, establish educational effectiveness or guarantee that every supported document behaves correctly in every environment.

Current Limitations and Future Work

StudyBridge AI is an evolving prototype. Future work includes stronger semantic retrieval, broader evaluation using more diverse learning materials and student profiles, improved assessment types, and support for richer multimodal content such as diagrams, scanned documents, and lecture videos.

We also aim to evaluate whether adaptive review and personalized recommendations measurably improve learning outcomes over time.

Our Vision

We want to make personalized learning more accessible to students who may not have access to private tutors or expensive educational technology. By connecting a student's own learning materials with grounded tutoring, adaptive practice, and visible progress, StudyBridge AI aims to make independent study more focused, transparent, and effective.

Built With

Share this project:

Updates

Submission history