Study Buddi

The inspiration behind Study Buddy was to create an interactive, AI-powered tool that helps students study efficiently by providing personalized quiz generation and answering questions from uploaded study materials, like PDFs.
The goal was to combine learning with engagement, offering an intelligent way to interact with content and get immediate feedback, which is especially useful in exam preparation or content review.
It was motivated by the need to address the problem of passive learning by turning study materials into an active, dynamic process.

Study Buddy allows users to upload PDFs (such as textbooks or lecture notes) and asks questions based on the content.
It generates 10 multiple-choice questions with real-time feedback on correctness and explains why a specific answer is correct or incorrect.
Users can also ask custom questions about the document, and the system retrieves relevant sections of the text to answer these questions using AI, enhancing comprehension.
It offers a visually engaging user experience with features like a typewriter effect for text and a streamlined interface.

The project was built using Streamlit for the front-end interface, allowing for a user-friendly web app experience.
LangChain was integrated to manage the workflow for question answering and quiz generation, especially to structure prompts and retrieve relevant document sections using FAISS vector storage.
Google Generative AI models were used to handle both the conversational aspect for question answering and to generate quizzes based on document content.
The PyPDF2 library was used to extract and process text from PDF documents, and FAISS was employed for similarity search to retrieve the most relevant content from the documents.
Various CSS customizations were added for enhanced UI, including animations for better visual engagement.

Handling large PDF files and breaking them into chunks that the model could process while keeping the context intact was a significant challenge.
Integrating different components like LangChain, FAISS, and the Google API to work smoothly together involved overcoming compatibility and embedding issues.
Managing API token limits and model performance was a challenge when it came to handling larger or more complex documents.
Ensuring that the quiz questions generated were diverse and challenging enough based on limited content input was another hurdle.

Successfully implemented a Retrieval-Augmented Generation (RAG) system that allows users to get specific, relevant answers from large PDFs.
Built a dynamic quiz generator that provides users with immediate feedback and explanations, adding to an interactive learning experience.
Integrated a sleek and engaging UI with custom animations and features like a typewriter effect, making the app more fun and engaging to use.
Managed to implement a system where users can process complex documents without much delay, thanks to efficient text chunking and retrieval.

Gained experience in building AI-powered applications with a combination of natural language processing, embeddings, and similarity search techniques.
Learned how to work with LangChain for prompt management and chaining tasks, which improved the ability to handle sophisticated workflows.
Improved understanding of creating seamless interactions between different components like Google's Generative AI, FAISS, and Streamlit to deliver a cohesive user experience.
Developed problem-solving skills, particularly in overcoming token limits, managing API responses, and integrating multiple systems into one tool.

Expanding to support more file formats (like Word docs or web page scraping) to widen the range of study materials.
Adding more customizable quiz features, like difficulty levels or topic-based quizzes, to cater to specific user needs.
Enhancing the user interface with more gamified features like streaks, badges, or study goals to increase user engagement.
Integrating speech-to-text functionality so that users can ask questions or interact with the app using voice commands.
Exploring advanced AI features like deeper document summarization or note generation based on uploaded files to enhance the learning experience even further.