Clarity

Chatbot
Concept Material
Question Detail
Dashboard
Ask Chat A Question With Context
Concept Overview
Slides Detail
Notes Detail

Inspiration

Our project was inspired by the challenges many of our fellow computer science students faced when working with course materials and practice problems. We noticed that students often struggle to effectively manage and interact with course materials spread across various PDFs and resources. This inspired us to create Clarity, a platform that would streamline access to learning materials while providing AI-integrated tools for better comprehension and problem-solving.

What it does

Clarity is a learning platform that:

Automatically organizes notes, slides, and questions into structured computer science concepts with an easy-to-navigate interface
Features an intelligent AI bot that can answer questions about concepts and practice problems based on the course material
Enables users to track their progress through practice problems of varying difficulty
Offers a clean, intuitive interface for accessing educational materials
Supports uploading and processing PDF files for course content

How we built it

We developed Clarity using a full-stack approach:

Backend:

Python with FastAPI for the AI-powered backend services
PDF processing with PyMuPDF for document analysis and extraction
Vector search capabilities using FAISS for efficient similarity matching
LlamaIndex for document indexing and semantic retrieval
Anthropic's Claude API for powering our intelligent AI assistant
VoyageAI integration for additional AI capabilities
Node.js/Express server for handling file uploads and content delivery
MongoDB with GridFS for storing PDF files (with fallback to in-memory storage)

Frontend:

React/Next.js for the user interface and component architecture
Tailwind CSS for responsive, modern styling
Custom PDF viewer component for seamless document viewing
Interactive chat interface for communicating with the AI assistant
State management for tracking user progress and interactions

Challenges we ran into

During development, we faced several significant challenges:

PDF viewer implementation: Integrating a reliable PDF viewer proved difficult due to browser compatibility issues and JavaScript library limitations
MongoDB connection problems: We encountered SSL/TLS errors when connecting to MongoDB Atlas
PDF processing: Handling binary PDF files, parsing content was technically challenging
Component integration: Ensuring smooth interaction between different parts of the application required careful state management

Accomplishments that we're proud of

Despite the challenges, we achieved several key successes:

Developed an intelligent AI assistant powered by Claude that provides contextual help for specific concepts and problems
Implemented document understanding capabilities using LlamaIndex and FAISS for semantic search
Created a natural conversational interface that makes complex topics more approachable
Built a fully functional learning platform with an intuitive, responsive interface
Created a flexible server architecture that adapts to connectivity issues
Developed a modular frontend design that can easily scale with additional features

What we learned

Methods for effective AI integration in educational applications
Techniques for document indexing and semantic search using vector embeddings
Prompt engineering techniques to enhance AI assistant capabilities
Approaches to contextual awareness in conversational interfaces
Best practices for integrating Python and Node.js in a full-stack application

What's next for Clarity

Looking forward, we plan to enhance Clarity in several ways:

Expand the AI assistant's capabilities with more specialized knowledge in computer science domains
Implement features for the AI to generate practice problems based on user's weak areas
Add memory and progression tracking to the AI assistant to provide more personalized learning paths
Improve document indexing with fine-tuned embeddings for better search relevance
Improve PDF annotation and note-taking features with AI-assisted summarization
Develop collaborative learning features where the AI can facilitate group discussions
Create a mobile app version for learning on the go with AI support
Expand the website to include many other disciplines