Plantey 🌱

Inspiration 🌟

"Do what you can, with what you have, where you are."
– Theodore Roosevelt

My inspiration for creating Plantey comes from the belief that we can make a difference by caring for the environment with the resources we already have. I’ve personally discovered that even a small garden, with plants specifically suited to our region and nurtured using what’s available to us, can have a meaningful impact. It’s this simple yet powerful idea that drives Plantey – small actions leading to big changes for a healthier planet.

What it does ⚡

• 🌍 Geolocation-Based Plant Curation
  Plantey fetches your exact location using the Geolocation API and uses AI to curate plants that thrive in your specific region.

• 🌿 Plant Categories
  The plants are categorized into:

  • 🌱 Herbs
  • 🌳 Trees
  • 🍎 Fruits & Vegetables
  • 🌸 Flowers
    This allows users to select plants based on their specific needs.

• 🌱 KnowPlants
  Understand the full details about any plant, provided by AI, including:

  • 🌸 Blooming Season
  • 💰 Net Cost for Growing
  • 🌺 Flower Color
  • 🌿 Growing Season
  • 💧 Watering Requirements
  • 🌵 Drought Tolerance
  • 🐞 Insects
  • 🦠 Diseases
  • 🌱 Soil Nature
    And more—basically, every piece of information you need to grow a plant successfully!

• 🌿 Growing Tips
  Access the Top 10 Gardening Tips to help you manage your garden and efficiently care for your plants.

How we built it 🛠️

• 🔺 Frontend and Backend are powered by React.
• 🔺 The plant recommendation system was created using Llama 70b versatile LLM with Groq LangChain.
• 🔺 Geolocation API was used to fetch the user's exact location, town, country, and pin code.
• 🔺 The Stability-Large-Turbo model was used for generating plant images.

Challenges we ran into 🔍

Integrating the AI model based on the use case into the app was the main challenge, and connecting the location input with the AI's input for generating the plant recommendation system was a huge task. 😤

Additionally, grabbing the user’s search input for a particular plant and incorporating it to analyze the entire details about that plant was also a tiresome task. 🧐

Incorporating the plant search inputs into Stability-Large-Turbo for generating the image was tough as well, but at the end of the day, I was able to resolve all the challenges and finally created a working model. 🎉 I feel so happy! 😊

Accomplishments that we're proud of 🥳

• 🎉 Find the location dynamically using the Geolocation API precisely.
• 🎉 Developed a plant recommendation system powered by Llama-70b-versatile LLM.
• 🎉 Integrated a fully AI-powered plant information system with AI-generated plant images.

What we learned 💡

• 💡 I learned a lot about integrating generative AI into applications
  Exploring how to incorporate generative AI into different app features, allowing for dynamic content generation based on user inputs and preferences.

• 🔧 Fine-tuning based on the user's needs
  Understanding the importance of customizing AI models to deliver more precise, personalized results for users, enhancing their experience and ensuring the system adapts to their specific needs.

What's next for Plantey 🚀

• 🌐 To deploy it to the web
  Deploying the app to a web platform to ensure it’s accessible to users globally, making the plant recommendation system available anytime, anywhere.

• 📍 Developing a location-based nearby plant shop identifier
  Integrating a feature that helps users find nearby plant shops based on their location, making it easier to purchase recommended plants.

• 🤖 Fine-tuning the LLM model for more precise output
  Enhancing the Llama-70b-versatile LLM to provide more accurate and tailored plant recommendations and detailed plant information based on user inputs.

Built With

• gps
• groq
• huggingface
• langchain
• llama
• react