Touch Grass

Priyanshu Sarkar — Sun, 11 Oct 2026 18:03:41 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass

What I Built

I built TouchGrass, an open-source AI-powered outdoor activity planner designed to help people spend less time on screens and more time exploring the real world.

It helps users discover outdoor places, generate activity plans, take on outdoor quests, and track their progress. The project combines a React and TypeScript frontend with a Python FastAPI backend, SQLite for data storage, OpenStreetMap data for place discovery, and local AI inference using Ollama with an open-weight model.

The goal is simple: use AI to encourage people to go outside rather than spend even more time online.

Demo

🌿 Live Demo: https://touchgrass-gsum.onrender.com/

💻 Source Code: https://github.com/priyanshu2734/touchgrass

Code

The complete source code, setup instructions, and contribution guidelines are available on GitHub:

https://github.com/priyanshu2734/touchgrass

Contributions and feedback are welcome!

How I Built It

I built TouchGrass using React, TypeScript, and Vite for the frontend, with Python and FastAPI powering the backend. SQLite handles local data storage, while OpenStreetMap services help discover nearby outdoor places.

For AI, the project is designed to use Ollama with the open-weight Gemma model, allowing inference to run locally instead of depending entirely on a proprietary, closed AI API.

This project brings together frontend development, backend APIs, external map data, local AI inference, and activity-planning features in one application.

Why Does Open Innovation Matter?

Open innovation makes it possible for developers to inspect, modify, and build upon the tools and models they use. By building TouchGrass around open-source technologies and an open-weight AI model, I can experiment with local inference, customize the application, and make the project accessible to other developers.