This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass
What I Built
I built an autonomous navigation and outdoor exploration assistant designed to bridge the gap between digital workflows and real-world environments. The tool helps users step away from their screens by dynamically generating optimized, nature-immersive outdoor routes, tracking location milestones in real time, and gamifying offline exploration. It is built for developers, students, and remote workers who spend long hours indoors and need a smart, frictionless nudge to get outside and "touch grass."
How I Built It
AI & Models: Utilized open-weight models running locally for lightweight, privacy-focused intent parsing and landmark recommendation.
Architecture: Built with React, TypeScript, and FastAPI, integrating spatial mapping libraries to handle real-time geolocation tracking and route adjustments on the fly.
Agent Harness: Developed custom contextual routing logic to prioritize scenic, low-traffic walking paths over standard vehicular roads.
Why Does Open Innovation Matter?
Open-source AI and transparent geospatial mapping make it possible to build localized, privacy-respecting navigation tools without relying on restrictive closed APIs. Having full control over the model weights and mapping layers allowed for offline-first capabilities, ensuring the app remains reliable even on remote trails where cellular data might drop.
My Agent Session
Prize Categories
Open-Source AI Challenge Week 1: Touch Grass
Top comments (0)