This is my submission for the Hacktoberfest Open-Source AI Challenge — Week 1: Touch Grass.
What I Built
We spend so much time looking at screens that sometimes we forget to look at what's around us.
A plant growing beside a road, a leaf with an unusual shape, or a tiny flower in a park can easily go unnoticed when we're busy scrolling. I wanted to build something that could make people a little more curious about the world outside.
That's how Jadi-Booti Solo came to life. 🌱
It's a nature exploration app designed to encourage people to step outside, notice plants around them, and record what they discover. You can take or upload a plant photo to try identifying it, explore plant information, keep a nature journal, and save little memories from your outdoor time. It also includes features such as voice notes and optional location-based memories.
The idea isn't to make people spend more time on another app. It's to give them a reason to put their phones down, go outside, and use technology to learn something about the real world.
It's for anyone who wants to reconnect with nature, whether that's a student taking a walk around campus, someone exploring a local park, or simply a person who wants to make their daily walk a little more interesting.
Demo
🌐 Live Demo:
The best way to experience the idea is to take a walk, find an interesting plant, and try exploring it with the app.
Code
💻 GitHub Repository:
Jadi-Booti Solo 🌿
An AI-powered nature exploration app and botanical memory journal built for the Touch Grass Hackathon Challenge.
Touch Grass Philosophy: The screen is the shortest part of the experience. Use open AI to help people step away from devices, discover living flora in parks and neighborhoods, complete physical outdoor missions, and record place-linked memories.
🌟 The 3 Core Pillars
1. AI Plant Identification (Open-Weight Vision)
- Users take a photo with their phone camera outdoors or upload a leaf/bark/flower specimen.
- Analyzed via Meta Llama 3.2 Vision (
llama-3.2-11b-vision-preview) hosted on high-speed Groq LPU inference. - Provides common name, botanical binomial, confidence score, leaf/bark visual characteristics, traditional ecological uses, and a clear safety uncertainty notice.
- Audio narration reads out plant stories and observation notes in natural Hindi/English speech.
2. Nature Quests (Outdoor Missions)
- Physical field missions that motivate users to explore the real world
- 🍃 The Tri-Leaf Hunter: Find…
The project includes a web-based frontend, an Express backend, plant data, and features for plant exploration and saving nature journal entries.
I'd love for others to explore the code, try the project, and contribute ideas that could make it more useful.
How I Built It
I built Jadi-Booti Solo as a web app with a focus on keeping the experience simple and accessible.
Here's the basic setup:
- Frontend: HTML, CSS, and JavaScript for the interface.
- Backend: Node.js and Express for API routes and app functionality.
- Plant data: A curated collection of plant information.
- AI features: The app is designed to support plant identification from images and speech-to-text for voice notes.
- Nature journal: Users can save observations and memories from their explorations.
- Progressive web app features: A web app structure that can make the experience feel more app-like on supported devices.
One of the challenges was getting the different parts to work together, especially when moving from local development to deployment. A feature can work on your own computer and still fail online if the frontend and backend aren't connected correctly. That taught me that building a project isn't just about making the interface look good; the underlying pieces have to work together too.
I'm still working on making the deployed experience reliable, particularly the AI-powered features. I want to be honest about that rather than pretend every part is already perfect.
Why Does Open Innovation Matter?
For a project like this, open innovation matters because nature is something everyone should be able to explore, learn about, and contribute to.
Open-source tools give developers a chance to experiment, learn from existing work, and build useful things without having to create every component from scratch. Open-weight AI models also offer more flexibility to developers who want to understand how their AI features work and explore different ways to run them.
That flexibility is important for a project like Jadi-Booti Solo. AI can help people learn about plants, but the experience should remain accessible and should encourage people to observe the real plant instead of treating an AI answer as unquestionable truth.
I'd also like the project to grow through community contributions: better plant information, more regional plants, improvements to accessibility, and new nature challenges. Someone who knows more about local plants than I do could help make the project better for everyone.
That's one of the things I appreciate about open source: you don't have to build everything alone.
Final Thoughts
Jadi-Booti Solo is a small project, but the idea behind it means a lot to me.
Technology doesn't always have to keep us indoors or glued to a screen. It can also help us become more curious about the places we live in.
If this app encourages even one person to take a walk, notice a plant they would normally ignore, or learn something new about their surroundings, then I think it's doing something worthwhile.
For now, my challenge to you is simple: put your phone away for a few minutes, step outside, and see what you notice. 🌿
You might be surprised by how much you've been walking past.
Top comments (1)
I built this because I realized how easy it is to ignore the world right outside my window. I'd love to know: what's your favorite local spot to unplug? Let me know in the comments!