🌿 TrailBuddy AI — Plan Less on Your Phone. Explore More Outside.

Rupam Gupta — Sat, 10 Oct 2026 17:51:23 +0000

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

What I Built

🌿 TrailBuddy AI — Plan Less on Your Phone. Explore More Outside.

TrailBuddy AI is an open-source, AI-powered outdoor activity planner designed to help people spend less time staring at screens and more time exploring the real world.

The idea is simple: technology should help us experience more of the world, not keep us glued to our devices.

Planning an outdoor activity can sometimes feel overwhelming. You might not know what to do, how to organize your available time, what to carry, or how to make the most of an outdoor experience. TrailBuddy AI brings these pieces together into one simple planning experience.

🌱 Key Features

🥾 1. Personalized Outdoor Activity Planning

Users can create plans for different outdoor activities:

Hiking and trail exploration
Walking and outdoor exploration
Running
Gardening
Birdwatching

📍 2. Customizable Preferences

Users can provide their preferred location, available duration, experience level, and interests to generate a plan tailored to their preferences.

🧭 3. Structured Outdoor Itineraries

TrailBuddy generates an organized itinerary that helps users decide what to do during their outdoor time instead of providing only a generic suggestion.

🎒 4. Packing Checklist

The packing checklist helps users prepare before leaving home. Users can review the items they need and track their preparation.

🍃 5. Nature Missions

TrailBuddy includes small nature-focused missions that encourage users to observe their surroundings, discover interesting details, and connect more intentionally with nature.

🛡️ 6. Safety Reminders

Generated plans include general safety recommendations to encourage responsible outdoor activities. Users should still verify current weather, trail conditions, access restrictions, and local regulations before leaving.

🖨️ 7. Print-Friendly Plans

Users can print their plans or save them as PDFs, making it easier to refer to their itinerary without repeatedly interacting with their phones.

🤖 8. Optional Local AI

TrailBuddy supports local AI inference using Ollama and the open-weight Qwen2.5 3B model. A built-in fallback planner allows the core planning experience to work even when local AI is unavailable.

📱 9. Simple User Interface

The React-based frontend provides an easy-to-use interface for entering preferences and viewing generated outdoor plans.

🎯 Who Is It For?

TrailBuddy AI is designed for:

People who want to build healthier outdoor habits.
Beginners who need help planning their first hike or outdoor activity.
Nature lovers interested in walking, birdwatching, and gardening.
Anyone who wants to make better use of their free time away from screens.

The goal is not to create another platform for endless scrolling. It is to help people plan an activity, prepare for it, and then put their phones away.

Demo

TrailBuddy AI can be run locally by following the setup instructions in the GitHub repository.

📸 Screenshots

1. TrailBuddy AI — Main Interface

2. Personalized Outdoor Plan

3. AI-Generated Outdoor Plan

Code

🌿 GitHub Repository: https://github.com/Rupam1210/Trail_BuddyAI

The repository contains the React frontend, Spring Boot backend, application configuration, and instructions for running the application locally.

Technology Stack

React.js + Vite: Frontend and development server.
Java: Backend programming language.
Spring Boot: REST API and application logic.
Maven: Backend dependency management.
Ollama: Local inference runtime.
Qwen2.5 3B: Open-weight language model.
REST API: Communication between the frontend and backend.

How the Application Works

The user enters outdoor activity preferences in the React frontend.
The frontend sends a request to the Spring Boot backend.
The backend attempts to generate a plan using the locally configured Ollama model.
If local AI is unavailable, the built-in deterministic fallback planner provides an alternative.
The generated plan is returned to the frontend for display.

The resulting plan can include an itinerary, packing checklist, nature missions, and safety recommendations.

API Endpoints

GET /api/health — Checks backend health and local AI availability.
POST /api/plan — Generates an outdoor activity plan based on the supplied preferences.

How I Built It

I wanted to combine a practical real-world use case with an accessible, open-source AI stack.

I built the frontend using React and Vite and used Java with Spring Boot for the backend logic. The frontend and backend communicate through a REST API using JSON.

For the AI component, I chose Qwen2.5 3B through Ollama as an optional local inference setup. This allows developers to experiment with an open-weight model without making a paid, hosted AI API a mandatory dependency.

I also included a deterministic fallback planner so that users can try the application's core functionality even if they have not installed or configured Ollama.

Challenges and Design Decisions

Making AI optional: Local AI setup can require additional downloads and system resources. The fallback planner makes the application easier to explore without requiring the model.

Keeping the application modular: Separating the React frontend from the Spring Boot backend makes it easier to improve the interface, planning logic, and AI integration independently.

Focusing on practical output: Rather than returning only a conversational response, TrailBuddy aims to provide an actionable plan with an itinerary, checklist, nature missions, and safety reminders.

Keeping the scope realistic: The current application does not fetch live weather, maps, trail conditions, park hours, or local regulations. These are possible future improvements.

Why Does Open Innovation Matter?

Open innovation matters because developers should be able to experiment with AI, learn how it works, and build useful applications without always depending on proprietary services.

For TrailBuddy AI, open innovation makes several things possible.

🔓 Freedom to Experiment

Using an open-weight model gives developers the opportunity to customize prompts, test different models, and improve the planning experience.

💻 Local Inference

When configured for local inference, prompts are processed by a model running on the user's machine rather than being sent to a hosted AI inference API. This offers more control over how prompts are processed, although it does not eliminate every possible privacy consideration.

💸 Reduced Dependency on Paid APIs

TrailBuddy does not require a paid AI API key for its core planning experience. Local AI inference uses the user's own computing resources after the model has been downloaded.

🌍 Community-Driven Improvement

Other developers can contribute new outdoor activities, improve fallback planning logic, enhance accessibility, refine the interface, or experiment with additional open-source models.

Most importantly, open innovation supports the idea behind TrailBuddy itself: technology should empower people to do more in the real world.

My Agent Session

I don't currently have a DevRelay agent session to share. This section can be updated with a session link or embedded session if one is recorded, following the challenge instructions.

TrailBuddy AI is built around the Touch Grass theme and includes optional local inference using an open-weight AI model.

🌳 Plan the adventure. Prepare for the outdoors. Then put the phone away and go live it.

I'd love to hear your feedback, suggestions, and contributions from the open-source community!

DEV Community: Rupam Gupta

🌿 TrailBuddy AI — Plan Less on Your Phone. Explore More Outside.

What I Built

🌿 TrailBuddy AI — Plan Less on Your Phone. Explore More Outside.

🌱 Key Features

🎯 Who Is It For?

Demo

📸 Screenshots

1. TrailBuddy AI — Main Interface

2. Personalized Outdoor Plan

3. AI-Generated Outdoor Plan

Code

Technology Stack

How the Application Works

API Endpoints

How I Built It

Challenges and Design Decisions

Why Does Open Innovation Matter?

My Agent Session