-
Simply upload a photo of your plant or a sick leaf, and let our advanced AI do the rest in seconds.
-
Explore a vibrant feed of plants shared by our users. Get inspired, connect, and showcase your own green friends!
-
Plan and visualize your perfect garden layout tailored to your specific space and plant requirements.
-
Set up automated email alerts for watering and maintenance so you never miss a beat.
-
Your personal plant hub. Easily manage your collection, track growth, and access saved schedules.
Inspiration
To be honest, our team used to be a group of notorious "plant killers." Our apartments were graveyards for ferns and monsteras. Whenever a plant got sick, we'd spend hours scrolling through traditional gardening forums, waiting anxiously for conflicting advice. We realized there had to be a better, faster way to get reliable botanical help. We were inspired to build Gardening Tips to give every plant parent a master botanist right in their pocket.
What it does
Gardening Tips is the ultimate AI companion for plant lovers. It features:
- Instant Identification: Recognize any plant, flower, or tree instantly with an incredible 98% AI accuracy.
- Smart Disease Diagnosis: Snap a photo of a sick leaf, and our AI instantly diagnoses the issue and provides an actionable treatment plan.
- Personalized Care Guides: Generates tailored advice on watering schedules, sunlight requirements, and soil types based on the exact plant species.
How we built it
We wanted a blazing-fast user experience, so we built the frontend using React and Next.js, styled with Tailwind CSS. The backend is powered by Node.js and PostgreSQL to securely store user plant profiles and schedules. For the "brain" of the app, we integrated a Python microservice using TensorFlow for computer vision (image classification for plant ID and disease detection), combined with the OpenAI API to generate natural, easy-to-understand care instructions.
Challenges we ran into
The biggest hurdle was achieving high accuracy for disease diagnosis. Plant symptoms can look very similar across different species, and lighting conditions in user photos vary wildly. We had to heavily preprocess the image data and fine-tune our computer vision models. Another challenge was prompt-engineering the LLM to ensure the gardening advice was practical and scientifically accurate, rather than generic.
Accomplishments that we're proud of
We are incredibly proud of hitting that 98% accuracy mark on plant identification. It feels like magic every time it works. We're also proud of taking complex AI architecture and abstracting it into a clean, minimalist, and friendly UI that even the most non-tech-savvy plant parent can use effortlessly.
What we learned
Technically, we leveled up our skills in integrating computer vision models with full-stack web applications. We also learned a massive amount about prompt optimization for domain-specific tasks. On a personal level? We learned a lot about botany! We actually know how to keep our monsteras alive now.
What's next for Gardening Tips
We're planning to build a community feature where users can swap plant cuttings locally. We are also exploring IoT integrations—imagine connecting smart soil moisture sensors directly to the app so the AI can alert you the exact moment your plant gets thirsty!
Log in or sign up for Devpost to join the conversation.