DEV Community

Shivansh Garg
Shivansh Garg

Posted on Fully Autonomous

I built Pocket Tutor, a local Gemma study buddy for a friend

#devchallenge #weekendchallenge #hf26challenge

Hacktoberfest Weekend Challenge: Build for a Friend Submission 🤝

This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend.

What I Built

I built Pocket Tutor, a study buddy for a friend who wanted studying to feel more visual and interactive.

You can start with a topic—even if you don’t have notes ready—or paste in your notes. Pocket Tutor uses them to create a short theory lesson, flashcards, and a four-question quiz. It also offers links to find videos and other study resources.

The feedback guiding the next version is clear: add videos, make practice more interactive, and offer more questions. The current version is a first step; the resource links open searches rather than curated videos, and the quiz is still short.

Demo

Screen Recording 2026-10-04 223401.mp4 - Google Drive

favicon drive.google.com

Code

GitHub repository: Pocket Tutor

Code

GitHub logo Shivansh1251 / pocket-tutor

A local-AI study buddy that turns notes or a topic into a short lesson, flashcards, and a quiz with optional study links.

Pocket Tutor

Pocket Tutor is a small, local-AI study buddy. Enter a topic for a beginner-friendly lesson, or paste class notes to make a guide grounded in that material. It adds key ideas, four flashcards, and a one-question-at-a-time quiz.

Why this is useful

The student can start even when they have no notes ready. A topic-only guide gives them a first explanation and practice questions; pasted notes keep the generated study set focused on the material they need to revise. The app can also open searches for videos and free reading when the student wants a visual explanation or another source.

Stack

  • Next.js App Router + TypeScript
  • Ollama for local inference
  • Gemma 3 1B (gemma3:1b) by default, chosen to run on computers with limited memory
  • No account, cloud inference API, or cloud database

Run locally

  1. Install Node.js 20.9 or newer and Ollama.

  2. Start Ollama and download Gemma:

    …
    Enter fullscreen mode Exit fullscreen mode
View on GitHub

How I Built It

Pocket Tutor uses Next.js for the app and its API route. The route sends the topic or notes to Ollama running locally, where Gemma generates structured content for the lesson, flashcards, and quiz. The app displays those as separate study steps.

The AI generation runs on the user’s computer through Ollama. The optional video and reading links open external searches, so those need an internet connection.

Why Does Open Innovation Matter?

Using local Gemma means study notes can stay on the learner’s computer instead of being sent to a hosted AI provider. After downloading the model, lesson generation doesn’t need a paid inference API. Running through Ollama also gives me room to change models and adapt the prompts as I improve the tutor.

Local inference uses the computer’s own memory and processing power, and generated explanations should still be checked against trusted course material.

Prize Categories

Best Use of Gemma

Top comments (0)