What I Built

I built Study Buddy, a personalized AI study assistant for my brother who usually understand stuff easily but forgets it quite quickly.

So instead of making him re-read everything , the study buddy helps him to get an overall review of his notes and it can make a small quiz for him as well and concise important points.

It can:

Explain concepts in simple language
Explain concepts using real-world analogies
Turn the material into concise, exam-ready answers
Generate quizzes to test understanding
Answer questions based primarily on the study material

Code

vidushi9746-cmyk / study-buddy

Study Buddy

A personalized AI study assistant built for the Hacktoberfest 2026 Build for a Friend challenge.

What it does

Study Buddy lets a student paste their notes, ask a question, and choose how they want the answer:

Explain simply
Give an analogy
Make an exam-ready answer
Generate a quiz

The project uses Gemma 4, an open-weight model, through the Gemini API.

Run locally

Install Node.js.
Clone/download this project.
Run:

npm install

Copy .env.example to .env.
Add your Google AI Studio API key:

GEMINI_API_KEY=your_key

Start:

npm start

Open http://localhost:3000.

Why open innovation matters

The app is built around an open-weight model rather than a closed proprietary model. This makes the model choice part of the project itself and gives the project a more transparent, adaptable AI foundation.

MVP scope

The intentionally small MVP focuses on the core learning loop:

student notes → personal question →…

View on GitHub

How I Built It

Study Buddy is built around Gemma 4, an open-weight AI model.

The application uses:

Node.js + Express for the backend

HTML, CSS and JavaScript for the frontend

Gemma 4 for generating explanations, analogies, exam answers and quizzes

PDF text extraction so students can upload their own study materia l

The AI receives the study material together with the question and the selected learning mode. The prompt tells the model to use the material provided as the primary source and to clearly say when the material does not contain enough information.

For Quiz Mode, the model is instructed to return structured questions that the frontend turns into an interactive quiz.

I intentionally kept the project small and focused. The goal was to make something my brother or anyone actually could actually use while studying rather than building unnecessary features around the AI.

Why Does Open Innovation Matter?

Open innovation mattered because the core of this project is the ability to build a focused application around an open-weight model rather than treating AI as a black-box feature that is completely controlled by a closed platform.

Using Gemma made it possible to explore how an open model can become part of a product designed around a specific person's needs.

It also makes the project easier to experiment with and extend. The AI isn't the entire product — it is the component that makes the personalized study experience possible.

For me, that is the most interesting part of open innovation: you can take an open model and build something much more specific and personal on top of it.

DEV Community

I built a Study Buddy for my brother who always forgets what he studied 10 mins ago ..... like literally 😭