Farm Intel — AI-Powered Agricultural Decision System

Farm Intel is a full-stack platform that provides data-driven insights for modern agriculture. It combines computer vision, environmental data, and rule-based analysis to assist in crop health monitoring, fertilizer planning, and decision-making.

---

Overview

The system is designed to help farmers and agronomists make informed decisions using real-time data and AI-assisted analysis.

It provides:

• Crop disease detection from images
• Soil-based crop recommendations
• Weather-aware decision support
• Fertilizer requirement calculations
• Interactive AI-based assistance

---

Features

Disease Detection

• Image-based crop analysis using vision models
• Identifies possible diseases, nutrient deficiencies, and stress factors

Crop Recommendation System

• Suggests suitable crops based on:

  • Soil pH
  • NPK values
  • Environmental conditions

Weather Integration

• Real-time weather data integration
• Forecast-based recommendations and alerts

Fertilizer Optimization

• Computes required NPK ratios
• Helps reduce overuse of fertilizers

AI Assistant

• Provides agricultural guidance and troubleshooting
• Handles user queries related to crops, soil, and environment

Dashboard

• Centralized interface displaying:

  • Weather data
  • System modules
  • Quick access tools

---

Tech Stack

Frontend

• React (SPA architecture)
• Vite (build tool)
• Tailwind CSS (styling)
• Framer Motion (animations)
• React Router (navigation)

Backend

• FastAPI (Python)
• Uvicorn (ASGI server)
• REST API architecture

AI & External Services

• OpenAI (vision + language models)
• OpenWeatherMap API (weather data)

---

System Workflow

1. User inputs data (image, soil values, or query)

2. Backend processes input through appropriate module:

   • Vision model → disease detection
   • Rule-based logic → crop/fertilizer recommendation
   • API → weather data

3. Results are returned via API

4. Frontend displays actionable insights

---

Installation

Prerequisites

• Node.js (v18+)
• Python (3.11+)
• API keys (OpenAI, OpenWeatherMap)

---

Clone the repository

git clone https://github.com/your-username/terra-intelligence.git
cd terra-intelligence

---

Backend setup

pip install -r requirements.txt
python -m uvicorn api.main:app --reload --port 8000

---

Frontend setup

npm install
npm run dev

---

Environment Variables

Create a .env file:

OPENAI_API_KEY=your_key_here
OPENWEATHERMAP_API_KEY=your_key_here
DEFAULT_CITY=Ernakulam
DEFAULT_COUNTRY=IN

---

API Documentation

Available at:

http://localhost:8000/docs

---

Project Structure

terra-intelligence/
├── api/
├── frontend/
├── models/
├── utils/
├── requirements.txt
└── README.md

---

Future Improvements

• Improve accuracy of disease detection models
• Add satellite or drone-based monitoring
• Offline support for low-connectivity areas
• Expand dataset for region-specific recommendations

---

Author

Joel Pradham AI/ML Engineer | Full-Stack Developer

---

License

MIT License