Your World Is More Interesting Than Your Feed.
This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass.
What if the things you noticed on a walk mattered more than the posts you scrolled past?
A strange-looking plant. An interesting rock. The sound of birds from a place you've never stopped to notice. Tiny details we usually walk past without a second thought.
That's the idea behind TRACE.
TRACE is a map-based exploration app that turns real-world observations into a personal, AI-assisted record of the world around you. It encourages you to step outside, pay attention, capture what you find, and slowly build a map of your own discoveries.
The goal isn't to spend more time in an app. It's to give you a reason to spend more time noticing the world outside it.
🌍 What I Built
TRACE combines real-world exploration, multimedia capture, local AI, and a little gamification to make curiosity a habit.
Here's what you can do with it:
- 🗺️ Trace Map: Explore observations on an interactive map and see the world through the things people notice.
- 📍 Trace Capture: Create observations with text, photos, or audio. Attach a location using GPS or place it manually on the map.
- 🧠 Local AI Processing: Turn captured observations into more structured records using a locally running AI model.
- 🌱 Sticker Garden: Earn daily stickers for qualifying observations, build exploration streaks, and collect sticker artwork as you keep exploring.
- ✨ A Personal Exploration Record: Revisit your saved traces and build a growing record of the places, sights, and sounds you've encountered.
The small details matter. A photo captures what you saw. Audio captures what you heard. A location gives that moment a place on the map. Together, they help turn an ordinary walk into something worth remembering.
🎥 Demo
Watch TRACE in action:
WhatsApp Video 2026-10-10 at 6.50.45 AM.mp4 - Google Drive
drive.google.com
In the demo, I'll walk through the map, create an observation using multimedia input, show the local AI processing flow, and finish with Sticker Garden and its daily reward system.
💻 Code
GitHub repository: https://github.com/shiriei/TRACE
TRACE is built around the idea that useful AI experiences don't always need to live in the cloud. The project brings together a map-based interface, real-world data capture, local inference, and a reward system designed around exploration.
Check out the repository to explore the implementation and follow the project's development.
🛠️ How I Built It
One of the central parts of TRACE is its use of local AI through LM Studio, running a 4-billion-parameter model on my machine.
Instead of sending every observation to a hosted AI service, TRACE connects to the locally running model for its AI processing workflow.
That choice comes with a trade-off: local inference can take a little time, depending on the hardware and input. But it also makes local execution a core part of the experience rather than an afterthought.
The rest of the application brings together an interactive map, text/photo/audio observation capture, saved traces, and the Sticker Garden reward system.
I wanted these pieces to feel connected. The map gives observations a place, multimedia capture preserves different kinds of experiences, AI helps structure the captured information, and daily rewards give users a reason to keep exploring.
🔓 Why Does Open Innovation Matter?
AI shouldn't be limited to applications that can afford a hosted API or send every request to an external service.
Open-weight models and local inference make it possible to experiment with AI in a different way: run models on your own hardware, build around them, and understand the practical trade-offs for yourself.
For TRACE, this meant being able to integrate a local model directly into an exploration workflow. It also meant working within the constraints of local inference, including processing time and available hardware.
Open innovation makes experimentation more accessible. Developers can build on available models, adapt them to their ideas, and create experiences that don't depend entirely on a closed AI service.
For me, TRACE is one small experiment in that direction: using local AI not just to generate text, but to support an activity that happens away from the screen.
🌱 The Bigger Idea
Gamification can easily become another reason to keep staring at a screen. I wanted to use it differently.
The map gives you a reason to notice. The capture tools help you preserve what you find. The AI helps organize your observations. Sticker Garden rewards consistency.
The real reward, though, is the world you're paying attention to.
Go outside. Notice more. Leave a trace.
🏆 Prize Categories
Best Use of Gemma
Use Gemma, Google's open-weight model, in building your project: run it locally, fine-tune it, or serve it through Google Cloud or another provider.
Built for the Hacktoberfest Open-Source AI Challenge — Week 1: Touch Grass.
Top comments (0)