Welcome to Study-Mate- AI — A simple, student-friendly entry point to personalized AI-powered learning.
Your Personalized Workspace — A unified interface for managing learning materials, interacting with the AI tutor, and accessing study tools.
Start Learning with Your Materials — Explore the StudyMate AI workspace and begin asking questions about your study material.
AI-Powered Learning Explanations— StudyMate AI helps students understand concepts through contextual explanations and source-based learning.

## Inspiration

Every student knows that feeling. You have PDFs, Word documents, lecture slides, your own notes all scattered everywhere. And when you actually need something, finding it takes longer than it should. So you turn to a general AI chatbot, get an explanation back, and then you're sitting there wondering does this even match what my lecturer taught? You can't verify it. You can't trust it fully.

That's exactly the problem we wanted to fix. That's where Study-Mate -AI came from.

The idea was simple but real what if students had an AI tutor that actually knew their coursework? Not a generic chatbot. Something that works from your own materials and helps you understand what's already in front of you.

What It Does

You upload your course documents, and from there you can ask questions, explore explanations, and interact with a tutor that uses your own learning materials as its context. We built two different ways to study with it:

Smart Tutor helps you understand concepts using whatever you've uploaded, and can also go broader when your documents don't cover something.
Grounded Study stays strictly within your uploaded materials so your answers are always connected to what your lecturer actually taught.

We also added source references and inspection tools so the AI isn't just giving you answers you have to blindly trust you can see where things are coming from.

The goal was always active learning. Not just generating answers, but helping students actually engage with their content.

How We Built It

We used React, TypeScript, Vite, and Google's Gemini AI. We started experimenting in Google AI Studio to feel out how the tutoring experience should work, then moved everything into GitHub as things progressed. We used VS Code throughout, with AI-assisted tools helping us move faster and debug along the way.

For the interface we drew inspiration from NotebookLM's clean design philosophy, but we didn't want to copy anything we wanted Study-Mate -AI to have its own identity built around students. Vercel handled deployment.

Challenges We Ran Into

Honestly, it wasn't a smooth ride and we're not going to pretend it was.

The hardest part was getting the AI to understand when to use uploaded materials and when to go general. During testing we found it was sometimes pulling from unrelated documents to answer questions. We had to go back and rework the retrieval and grounding logic to make it actually reliable.

Then came the frontend redesign. The new interface looked so much better — but integrating it broke things that were already working. Features got disconnected, some were replaced with placeholders, and we had to dig through API errors, restore real functionality, and reconnect data without losing the design improvements we'd made.

We also hit Gemini API timeouts, usage limits, and deployment configuration problems all under hackathon time pressure. Every decision had weight.

The lesson that hit hardest: a polished interface means nothing if what's underneath it doesn't work.

Accomplishments We're Proud Of

We didn't stop at an idea or a presentation. We built something that actually works, and we kept pushing even when things broke close to the deadline:

Built a full document-based AI tutoring experience
Implemented two distinct learning modes Smart Tutor and Grounded Study
Developed source-grounding and citation inspection functionality
Designed a clean, student-focused learning workspace
Collaborated through GitHub while troubleshooting real integration problems
Learned how to prepare and deploy a web application end to end

What We Learned

This wasn't just about connecting an AI model to a website. We learned retrieval and source grounding, frontend and backend integration, API error handling, version control, interface design, and deployment all in one go.

AI-assisted development helped us move faster, but it never replaced the need to actually think, debug, test, and make judgment calls.

Building with AI is not just about generating code. It's about understanding the problem, evaluating what comes back, and taking responsibility for what you put out into the world.

What's Next for Study-Mate- AI

We want to keep making Study-Mate- AI more reliable and more accessible better source retrieval, stronger topic organization, smoother assessments, and a mobile experience that works even on limited internet, because not every student has the best connection or device.

Eventually we want it to help students identify exactly where they're struggling and focus their revision there.

The long-term vision is real to us quality AI-assisted learning for every student, especially those who can't afford expensive tutoring.

Study-Mate- AI is our step in that direction. One student at a time.