FitBuddy Outdoor Quest 🌱 — Local AI That Gets You Outside

What I Built

For Hacktoberfest Week 1's Touch Grass challenge, I extended my existing project, FitBuddy, with an AI-powered Outdoor Quest feature.

FitBuddy was originally built to help my friend Atharva develop healthier habits through small, gradual changes. With Outdoor Quest, I wanted to connect AI with something beyond the screen: spending a little more time outdoors.

The feature generates simple, beginner-friendly outdoor missions, such as taking a short walk, noticing different plants, or observing the world around you.

After completing a quest, users can mark it as done. The app keeps a completion count in browser storage.

The idea is simple: use technology to encourage people to spend more time away from technology. 🌳

Demo

The project currently runs locally on my computer.

I have included screenshots of the Outdoor Quest interface, AI-generated missions, and the completion message.

How I Built It

Tech stack:

HTML

CSS

JavaScript

Node.js and Express

Ollama

Gemma 3:1B

Browser LocalStorage

The frontend sends a request to my Node.js backend, which calls the locally running Gemma 3:1B model through Ollama. The generated quest is then displayed on the webpage.

I also added a completion button that stores the number of completed quests using LocalStorage.

Building this feature helped me understand how a locally running open-weight AI model can power an interactive web application.

Why Does Open Innovation Matter?

Open innovation makes AI experimentation more accessible to students and independent developers.

By using an open-weight model with Ollama, I could experiment with AI locally without depending on a paid, hosted AI API. It gave me an opportunity to learn by building something practical with tools available to individual developers.

I believe AI should not only help us do more on our screens. It can also encourage us to step away, explore our surroundings, and build healthier habits.

This project is a small experiment in that direction. 🌱

Prize Categories

Submitted for the Hacktoberfest Week 1 challenge: Touch Grass.