🌸 Flower Recommender

A garden planning tool that helps you find the perfect flowers for your garden based on climate, sun exposure, watering needs, and other growing conditions.

Features

• 🔍 Flower search based on multiple criteria:
  • Climate zones
  • Sun exposure requirements
  • Watering needs
  • Soil type preferences
  • Blooming seasons
  • Hardiness zones
• 🎨 Color-coded flower cards based on flower descriptions
• 📊 Detailed growing information for each flower
• 🌿 Database of 75+ flowers
• 🎯 Semantic search powered by Pinecone Database
• 🤖 Integrated vector embedding using multilingual-e5-large model hosted by Pinecone

Tech Stack

• Next.js 14 with App Router
• TypeScript
• Tailwind CSS
• Pinecone Database
• Environment: Node.js 18+

Prerequisites

1. Node.js 18 or higher
2. A Pinecone account
3. npm or yarn package manager

Setup

1. Clone the repository:

   git clone https://github.com/yourusername/flower-recommender.git
   cd flower-recommender
   

2. Install dependencies:

   npm install
   # or
   yarn install
   

3. Create a .env file in the root directory with your Pinecone credentials:

   PINECONE_API_KEY=your_api_key_here
   PINECONE_INDEX_NAME=flower-recommender
   

4. Create the Pinecone index:

   npx ts-node scripts/create-index.ts
   

   This script will:

   • Check if the index already exists
   • Create a new index with integrated embedding using the multilingual-e5-large model
   • Configure automatic text embedding for flower descriptions
   • Wait for the index to be ready
   • Handle any errors during creation

   The index is configured with:

   • AWS cloud deployment in us-west-2
   • Multilingual-e5-large embedding model
   • Automatic text embedding for the 'description' field
   • Serverless architecture for automatic scaling

5. Seed the database with flower data:

   npx ts-node scripts/seed-flowers.ts
   

6. Start the development server:

   npm run dev
   # or
   yarn dev
   

7. Open http://localhost:3000 in your browser

License

MIT