# 🌿 PlantCare AI — "Less Scrolling. More Growing."
I'm building PlantCare AI, an AI-powered plant-care assistant that helps people understand potential plant health problems using plant images, observed symptoms, and practical care recommendations.
PlantCare AI combines a user-friendly interface with AI-powered analysis using Google Gemini to help users interpret visible plant symptoms, explore possible causes, and discover appropriate care steps.
This project is my submission for the Hacktoberfest Open-Source AI Challenge — Week 1.
🌱 What I Built
PlantCare AI is designed to make plant-health information more accessible to plant owners, gardening beginners, and anyone who wants to take better care of their plants.
Key Features
- 📷 AI Plant Doctor: Upload a plant image to start a plant-health assessment.
- 🔎 Symptom Selection: Identify visible symptoms such as yellowing leaves, brown or crispy leaf tips, wilting, dark blotches, and stunted growth.
- 🤖 Google Gemini Integration: Use Google's Gemini AI capabilities to support plant-health analysis and generate helpful responses.
- 🧠 AI-Generated Diagnosis: Present possible diagnoses, severity estimates, confidence information, potential causes, and recommended care actions when provided by the analysis workflow.
- 💧 Personalized Care Recommendations: Get practical guidance related to watering, sunlight, soil and roots, and potential pest problems.
- 🌿 Plant Selection: Provide information about the affected plant to add context to the assessment.
- 🗂️ Assessment History: Save and revisit previous plant-health assessments.
- 📱 Responsive User Interface: Access the application through a clean, responsive interface designed for an approachable plant-care experience.
🔄 How It Works
- Select the affected plant.
- Upload a photo of the plant.
- Select visible symptoms and optionally provide additional notes.
- Submit the information for AI-powered analysis.
- Review the possible diagnosis, contributing factors, and recommended care steps.
- Save the assessment and revisit it later.
PlantCare AI is intended to provide preliminary guidance rather than a guaranteed diagnosis. AI-generated results can be inaccurate, so users should consider the plant's growing conditions and consult a horticultural expert when necessary.
🛠️ How I Built It
Frontend — React
React powers the user interface, including plant selection, image uploads, symptom selection, and diagnosis results.
Backend — FastAPI
FastAPI provides the backend API and coordinates application workflows, including plant-health assessment requests.
AI — Google Gemini
Google Gemini provides the generative AI capabilities used for plant-health analysis and care guidance. The integration allows the application to use Google's AI models to interpret submitted information and generate helpful, natural-language responses.
Assessment Management
The application includes assessment-history functionality so users can save and revisit previous plant-health results.
Testing — TestSprite
I used TestSprite to create and run tests for important application workflows, including navigation, authentication, plant-image and form validation, diagnosis, and assessment history.
Challenges and Learning
Building PlantCare AI involved working through AI integration, model compatibility, inference responses, validation, and application reliability.
This experience reinforced an important lesson: building a useful AI application requires more than integrating a model. It also requires reliable backend workflows, meaningful error handling, testing, and an interface that helps users understand the results.
🚀 Live Demo
Try the deployed application:
https://plantcare-ai-six.vercel.app/
💻 Source Code
GitHub Repository:
https://github.com/yashbora18/Plantcare-ai
The project separates the frontend, backend API, and AI-related responsibilities to support maintainability and future improvements.
Contributions, feedback, and suggestions are welcome!
🌍 Why Open Innovation Matters
Plant-health problems can be difficult to identify, particularly for people who are new to gardening. An AI-assisted application can help users understand possible causes, explore appropriate care options, and decide what to investigate next.
By combining open-source application development with Google's Gemini AI capabilities, PlantCare AI explores how modern AI tools can be applied to an everyday problem.
The project also highlights the importance of responsible AI development: presenting results clearly, acknowledging uncertainty, validating inputs, and helping users make informed decisions rather than treating AI-generated answers as unquestionable facts.
🤖 AI-Assisted Development
I used an AI coding assistant to support implementation, debugging, and testing workflows during development.
The work included improving the plant-diagnosis workflow, troubleshooting integration issues, validating application behavior, and refining the overall user experience.
🏆 Challenge Categories
I'll select the applicable prize categories from those officially listed on the challenge submission page. I will not claim eligibility for a category unless it matches the published requirements.
🔮 What's Next?
- Improve diagnosis quality across a broader range of plants and symptoms.
- Expand plant-specific care recommendations.
- Improve error handling for failed or incomplete AI responses.
- Continue testing image uploads, validation, authentication, and assessment history.
- Enhance the accessibility and responsiveness of the user interface.
- Continue improving the deployed application based on feedback.
🌿 Final Thoughts
Building PlantCare AI has been a valuable opportunity to explore Google Gemini integration, full-stack development, AI-assisted problem-solving, and testing an AI-powered application.
My goal is to make plant-care guidance easier to access while continuing to improve the application's usefulness, reliability, and user experience.
I'd love to hear your feedback, suggestions, and contributions!
Top comments (0)