Personal Study Guide and Test Generator

• 

  Home Page

💡 Inspiration

We were inspired by the growing challenge students face in filtering trustworthy study content and preparing for tests efficiently. Most AI-powered tools often hallucinate facts, which is risky for academic learning. We wanted to create a solution that not only generates questions but ensures they're rooted in verified, human-curated material. This led to the idea of combining Large Language Models (LLMs) with Retrieval-Augmented Generation (RAG) to build a reliable and intelligent study companion.

🧠 What it does

The Personal Study Guide and Test Generator is a web-based platform that:

• Allows students to upload study material (PDFs, images, notes)
• Extracts and classifies key concepts and image-caption pairs
• Uses RAG and LLMs to generate context-aware, accurate test questions
• Displays interactive dashboards showing progress and weak areas
• Helps users self-evaluate with reliable, source-backed content

🛠️ How we built it

• Frontend: Built with React, HTML, and Tailwind CSS for responsive design
  
• Backend: Implemented using Python (FastAPI & Flask) to handle logic and API routes
  
• Document Processing: Parsed PDFs/images using OCR and structured extraction
  
• Database: Used MongoDB and PostgreSQL to manage content and user analytics
  
• LLM Integration: Connected HuggingFace models with a custom RAG pipeline to generate grounded questions
  

🚧 Challenges we ran into

• Handling various PDF/image formats and extracting clean, structured content
  
• Preventing hallucination in LLM-generated outputs — required careful tuning of the RAG pipeline
  
• Managing performance and latency when generating questions from large documents
  
• Designing a user interface that’s both informative and intuitive for students and educators
  

🏆 Accomplishments that we're proud of

• Successfully integrated a working RAG pipeline to eliminate hallucinated outputs
  
• Built a full-stack prototype capable of parsing real educational documents
  
• Delivered high-quality, relevant MCQs aligned with uploaded study content
  
• Created a foundation for personalized learning analytics through dashboards
  

📚 What we learned

• Practical implementation of Retrieval-Augmented Generation in educational tools
  
• Best practices in OCR and text/image parsing for academic content
  
• Structuring LLM prompts and embeddings to ensure context relevance
  
• Full-stack development from frontend UI to backend API and database operations
  

🔮 What's next for Personal Study Guide and Test Generator

• Add support for handwritten notes and textbook scans via advanced OCR
  
• Enable collaborative study groups and shared test sets
  
• Integrate with LMS platforms (e.g., Moodle, Google Classroom)
  
• Launch mobile support for on-the-go test generation and revision
  

Built With

• css
• embedding
• fastapi
• flask
• groq
• huggingface
• javascript
• languages-html
• llm
• mongodb
• postgresql
• python
• rag
• react
• restful
• transformer
• typescript
• vite
• webpack