# AI Garden Buddy: A Local AI That Helps You Touch Grass > Source: > Published: 2026-10-05 23:03:54+00:00 This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass **What I Built** I built AI Garden Buddy, a simple gardening assistant designed to help people spend less time on screens and more time outside. The idea is simple: instead of using AI just to answer questions, I wanted AI to give people a reason to step away from the computer and actually do something outdoors. AI Garden Buddy helps users create a simple gardening plan based on their location, season, available sunlight, garden space, and experience level. **It can suggest:** Plants that may be suitable for the user's garden Weekly gardening activities Watering and care suggestions Simple outdoor gardening missions A practical plan that the user can take outside For example, instead of spending 30 minutes chatting with an AI about gardening, the application can give you a task like: "Spend 20 minutes outside today. Prepare a small area for planting and check the soil moisture." The goal is to make the screen the starting point, not the destination. **Code** GitHub Repository: [https://github.com/360abhimanyu/AI-Garden-Buddy](https://github.com/360abhimanyu/AI-Garden-Buddy) The project is intentionally kept simple so that other developers can understand it, run it locally, and experiment with the AI component. **How I Built It** I built AI Garden Buddy using a lightweight web application with a local AI model. The main open-weight AI model I used is Qwen2.5 1.5B, running locally through Ollama. The basic flow looks like this: The application sends the gardening information to the locally running model, which generates recommendations and outdoor activities. The important part is that the AI doesn't need to be a large cloud service to be useful. A small open-weight model can handle the core experience while running on the user's own machine. I also added a fallback experience so that the application can still provide useful gardening suggestions when the local model isn't available. **Why Does Open Innovation Matter?** Open innovation matters because it gives developers more control over how AI is used. For this project, I wanted to avoid making another application that depends completely on a closed AI API. Using an open-weight model and local inference means the project can be experimented with without sending every gardening request to a third-party AI service. It also makes the architecture easier to change. I can experiment with different models, change the prompts, modify the application logic, or eventually fine-tune the model for gardening-related tasks. There is also an important privacy benefit. A user can run the AI locally instead of sending their information to a remote AI service. For me, the biggest benefit is experimentation. OpenAI makes it possible to take an idea, run it locally, understand how the pieces work, and change them without being locked into one provider. My Agent Session I used AI-assisted development to build this project and experimented with the application's structure, user experience, and open-source AI integration. **Agent Session**: I used AI-assisted development to help design and build AI Garden Buddy, including the application structure, user experience, and integration with the local open-weight AI model. I did not use a DevRelay session for this project, so there is no session link to share. **Prize Categories** **Final Thought** AI doesn't always have to keep us in front of a screen. Sometimes the best AI experience is one that gives you a useful answer and then tells you to close the laptop. That's what I wanted to build with AI Garden Buddy. Get the gardening plan. Go outside. Touch grass.