🌾 Eco-Farm Assist: Empowering Small & Marginal Farmers with AI

Eco-Farm Assist is an AI-driven agricultural advisory platform designed to empower India's small and marginal farmers. By combining the power of satellite data, weather APIs, and scalable ML models, it provides personalized and real-time advice on crop planning, fertilizer use, and disease detection.

🚜 Bridging the gap between technology and grassroot agriculture through intelligent, location-aware recommendations.

---

📌 Why Eco-Farm Assist?

• Traditional advisory systems provide generic, static advice that ignores location, seasonality, and soil type.
• Farmers lack real-time, personalized insights to make smart decisions.
• There is low adoption of AI and satellite data in rural agriculture.

Eco-Farm Assist solves this by delivering:

• 📍 Location-specific crop recommendations using Google Earth Engine
• 🌦️ Real-time weather-aware suggestions via weather APIs
• 🧠 Scalable ML models for intelligent farming decisions
• Early Detection and Optimal treatment of pests

---

🛠️ Tech Stack

Eco-Farm Assist is powered by a modern, scalable and modular tech stack:

🌐 Frontend

• Next.js 15 – Full-stack React framework for server-side rendering and routing
• Tailwind CSS + tailwind-merge + tailwindcss-animate – Utility-first styling with animation support
• Lucide React, Font Awesome, Geist UI – Icon & UI libraries for rich visual experience
• Framer Motion – Fluid animations for engaging UX
• React Hook Form + Zod + @hookform/resolvers – Robust form handling and validation
• Embla Carousel, react-scroll, usehooks-ts – Enhanced UI interactions and hooks

🔐 Authentication & Security

• NextAuth.js – Secure user authentication and session management
• JWT + bcryptjs – Token-based authorization and password hashing
• dotenv – Secure environment variable management

🌱 Backend & APIs

• Google Earth Engine – Satellite-based geospatial intelligence
• Google Generative AI (Gemini) – Smart query answering via NLP
• Vertex AI – Deployed ML models for crop, fertilizer, and disease prediction
• Mongoose + MongoDB – Scalable NoSQL database
• Nodemailer + React Email – Transactional emails for sign-up and verification

🧰 Developer Tools

• ESLint + Prettier – Code linting and formatting
• TypeScript – Type-safe development
• Axios – Promise-based HTTP client for all API communication
• PostCSS + Autoprefixer, clsx, glob – Utility libraries for modern development

---

📡 API Structure

Location: src/app/api/

├───auth
│   └───[...nextauth]
├───change-password
├───chatbot
├───confirm-password-validation
├───crop-prediction
├───disease-prediction
├───fertilizer-recommendation
├───geocode
├───get-latest-advice
├───password-validation
├───sign-up
├───soil-data
├───verify-code
├───weather-data

Key API Endpoints

• POST /api/sign-up – Register a new farmer
• POST /api/auth/[...nextauth] – Login and authentication
• POST /api/crop-prediction – Get crop recommendations based on soil & weather
• POST /api/fertilizer-recommendation – AI-based fertilizer suggestions
• POST /api/disease-prediction – Image-based crop disease detection
• GET /api/soil-data – Fetch soil pH, NPK, and moisture using GEE
• GET /api/weather-data – Real-time weather forecast for user location
• POST /api/chatbot – Smart conversational chatbot using Gemini
• GET /api/get-latest-advice – Daily farming tips personalized for user

---

🚀 Deployment

• Frontend: Render
• Backend: Same monorepo deployment on Render with serverless architecture
• GitHub: https://github.com/Sudhanshu-NITR/EcoFarm-Assist
• ML Repository: https://github.com/swaraj123822/EcoFarmAssist-model
• Demo Video: YouTube Link

---

🧪 Getting Started

📦 Prerequisites

• Node.js (v18+)
• npm or yarn
• .env file with required secrets

📂 Installation

git clone https://github.com/Sudhanshu-NITR/EcoFarm-Assist.git
cd eco-farm-assist
npm install

🚀 Running the App

npm run dev

🔐 Required .env Variables

# Database
MONGODB_URI=your_mongodb_connection_uri

# Email (Transactional Email via Brevo)
BREVO_EMAIL=your_email@example.com
BREVO_SMTP_KEY=your_brevo_smtp_api_key

# Authentication
NEXTAUTH_SECRET=your_nextauth_secret

# Google AI (Gemini + Vertex AI)
GEMINI_API_KEY=your_google_gemini_api_key
VERTEX_ENDPOINT_ID_CROP_REC=your_vertex_ai_crop_endpoint_id
VERTEX_ENDPOINT_ID_FER_REC=your_vertex_ai_fertilizer_endpoint_id

# Google Maps + Weather
NEXT_PUBLIC_GOOGLE_MAPS_API_KEY=your_google_maps_browser_api_key
NEXT_PUBLIC_OPENWEATHERMAP_API_KEY=your_openweathermap_api_key

# Google Cloud & Earth Engine
GCP_PROJECT_ID=your_gcp_project_id
GCP_REGION=your_gcp_region
GOOGLE_APPLICATION_CREDENTIALS_BASE64=base64_encoded_service_account_json
EARTH_ENGINE_SERVICE_KEY=your_earth_engine_service_key

---

📸 Screenshots

Landing Page

Crop Recommendation

Disease Detection

Fertilizer Recommendation

Weather Insights Page

---

🔭 Future Roadmap

• 🌍 Regional language support with dynamic translation
• 📱 Progressive Web App (PWA) version
• 🧠 Advanced pest detection model integration
• 📈 Market price prediction module
• 📡 IoT sensor integration for real-time field data

---

🤝 Team

Team Name: Red Bull Coders
Team Lead: Sudhanshu Kadam
Team Members:

• Swaraj Jamkar swaraj123822
• Sakshi Bihani sakshib19

---

“Farming is not just a profession, it's a responsibility. Eco-Farm Assist is our small step towards making it smarter, easier, and more rewarding.”