CampusCopilot：面向學生的個人 AI 開發與學習工作空間

CampusCopilot — Your Personal AI Developer Campus 🚀

Submission for the Hacktoberfest Weekend Challenge: Build for a Friend

What I Built

I built CampusCopilot, a personal AI learning workspace designed to help a college student manage DSA practice, coding mistakes, study materials, assignments, and exams in one place.

The idea came from a simple problem: students often use different tools for notes, coding practice, planning, and AI assistance, but these tools don't connect their learning progress.

CampusCopilot brings these activities together in one connected workflow.

Its key features include:

• 📚 StudyBuddy: Ask questions about uploaded study materials, learn concepts step by step, and practice with quizzes.
• 💻 CodeExplain: Understand coding errors through structured explanations covering what happened, why it happened, how to fix it, and what to learn next.
• 📅 Campus Planner: Organize assignments, exams, tasks, and study schedules.
• 📊 Learning Profile: Track learning activity, strengths, and weak areas.
• 🎯 Next Best Move: Get recommendations for what to focus on next.
• 🤖 Local AI: Connect to Llama 3.1 8B through Ollama for local inference.

The goal is not simply to give students answers. It's to help them understand their mistakes and decide what to do next.

Demo

🌐 Live Application: https://campus-copilot-flax.vercel.app/

🎥 Video Demo: ADD YOUR YOUTUBE VIDEO LINK HERE

💻 GitHub Repository: https://github.com/iam-ayushraj05/campus-copilot

The deployed application can be explored in Demo Mode. Local AI inference requires a compatible local Ollama setup.

How I Built It

CampusCopilot uses React, TypeScript, and Vite for the frontend, with a modular AI service layer that separates the application from its AI provider.

For local AI, it supports Llama 3.1 8B through Ollama.

StudyBuddy uses a retrieval pipeline based on document text extraction, chunking, TF-IDF, and cosine similarity to find relevant context from uploaded study materials.

The application also uses a shared state stored in browser local storage to keep learning activity and application data connected across its different sections.

The central learning workflow is:

Study → Practice → Identify Weaknesses → Recommend Next Steps → Improve

Why Open Innovation Matters

Open innovation matters because students should have more control over how AI processes their learning materials and coding context.

By supporting an open-weight model through Ollama, CampusCopilot provides a local inference option instead of requiring every interaction to use a closed cloud AI API.

This approach offers several benefits:

• More control over personal study materials and code.
• The ability to experiment with local AI models.
• Less dependence on a single cloud AI provider.
• Opportunities to customize the AI experience.
• The possibility of running inference locally without sending every prompt to a remote model provider.

Local inference does require a compatible machine and a configured model. The deployed demo also provides a way to explore the application without setting up local AI.

What Makes CampusCopilot Different?

CampusCopilot connects learning tools instead of treating them as separate features.

A student can study a topic, test their understanding, identify a weak area, practice coding, and use the planner to organize the next study session.

The bigger idea is that AI should help students make progress, not just generate answers.

Testing and Improvements

I tested the main application routes and interactions, added loading and empty states, improved error handling, and checked the production build.

The project was built iteratively, with testing and debugging throughout development.

Final Thoughts

CampusCopilot started with a simple question:

What if an AI learning assistant could help a student understand not only what they got wrong, but also what they should do next?

That question became a connected learning workspace combining study assistance, coding explanations, planning, and personalized recommendations.

My goal is to make learning more organized, understandable, and actionable for students.

Study smarter. Understand your code. Know what to do next. 🚀

AI Disclosure

AI tools were used for coding assistance, debugging, iteration, and documentation during development. Product direction, feature requirements, architecture decisions, testing, and final implementation decisions were guided by the project goals.