🌾 FarmAssist AI

AI-powered farm advisor for smallholder farmers worldwide

Built for the Amazon Nova AI Hackathon 2026
Category: Social Impact · Agriculture · AI for Good

🚀 Live Demo · 📹 Watch Demo Video · 🐛 Report Bug

---

🌍 The Problem

Over 33 million smallholder farmers across Africa and the developing world make life-changing decisions every single day — what to plant, how to treat sick crops, when to sell, where to buy supplies, with zero access to expert agricultural guidance.

A bad decision doesn't just mean a bad harvest. It means a family goes hungry. It means a child doesn't go to school. It means a livelihood is lost.

FarmAssist changes that.

---

💡 The Solution

FarmAssist is an AI-powered farm advisor that puts expert agricultural knowledge in every farmer's pocket — for free, in their language, available 24/7.

Powered by Amazon Nova through AWS Bedrock, FarmAssist understands natural language questions about crops, pests, weather, market prices, and more — and responds with clear, practical, actionable advice tailored to African and global farming contexts.

---

✨ Key Features

Feature	Description
🌱 Crop Planning	Personalized planting advice based on season, location & soil type
🐛 Pest & Disease Detection	Describe symptoms and get instant diagnosis + treatment plan
🌦️ Weather Guidance	Rainy and dry season farming tips tailored to your region
💰 Market Intelligence	Know which crops are most profitable right now
🧪 Fertilizer Guide	Exactly what to apply, when, and how much
🏪 Supply Locator	Find seeds, fertilizers, and equipment near you
🗣️ Voice Input	Speak your question — no typing needed
🌍 Multi-language	English, Yoruba, Hausa, Igbo, Français, Swahili
💬 Chat History	All conversations saved and accessible from the sidebar
📱 Fully Responsive	Works beautifully on mobile, tablet, and desktop

---

🎬 Demo

The app opens with a clean, professional splash screen before revealing a stunning chat interface packed with warm earthy greens and premium typography.

Try asking:

• "What crops should I plant this rainy season in Nigeria?"
• "I see yellow spots on my cassava leaves — what is it?"
• "What is the best fertilizer for maize and when should I apply it?"
• "Which crops are selling for the best prices right now?"

---

🛠️ Tech Stack

Frontend        →  Next.js 14 (App Router) + TypeScript + Tailwind CSS
AI Model        →  Amazon Nova Lite via AWS Bedrock
Rendering       →  react-markdown for rich formatted responses
Voice Input     →  Web Speech API (Chrome/Edge)
Icons           →  lucide-react
Fonts           →  Cabinet Grotesk + Satoshi + JetBrains Mono
Deployment      →  Vercel

---

🚀 Getting Started

Prerequisites

• Node.js 18+
• AWS Account with Bedrock access
• Amazon Nova model enabled in AWS Bedrock

1. Clone the repository

git clone https://github.com/your-username/farmassist-ai.git
cd farmassist-ai

2. Install dependencies

npm install

3. Configure environment variables

Create a .env.local file in the root directory:

AWS_ACCESS_KEY_ID=your-access-key-id
AWS_SECRET_ACCESS_KEY=your-secret-access-key

⚠️ Never commit your .env.local file. It is already in .gitignore.

4. Run the development server

npm run dev

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

---

📁 Project Structure

farmassist-ai/
├── app/
│   ├── page.tsx                    # Entry point
│   ├── layout.tsx                  # Root layout
│   ├── globals.css                 # Global styles & animations
│   ├── components/
│   │   └── FarmAssist.tsx          # Main UI component
│   └── api/
│       └── chat/
│           └── route.ts            # Amazon Nova API route
├── public/
├── .env.local                      # API keys (never commit)
├── tailwind.config.ts
└── README.md

---

🤖 How the AI Works

Every message is sent to Amazon Nova Lite via AWS Bedrock with a carefully crafted system prompt that instructs the model to:

• Act as an expert agricultural advisor for African farmers
• Give practical, concise, farmer-friendly advice
• Respond appropriately based on the selected language
• Structure responses with clear headings, lists, and actionable steps

User message → Next.js API Route → AWS Bedrock (Amazon Nova) → Formatted response → React Markdown renderer → Beautiful UI

---

🌱 Impact

FarmAssist is designed to directly address UN Sustainable Development Goal 2: Zero Hunger by empowering smallholder farmers with knowledge that:

• 📉 Reduces crop losses from preventable pest damage and poor planting decisions
• 📈 Increases income by connecting farmers to the best market opportunities
• 🍽️ Improves food security at the household and community level
• 🎓 Democratizes agricultural expertise previously only available to large commercial farms

---

🏆 Hackathon

This project was built for the Amazon Nova AI Hackathon 2026

Why Amazon Nova?

• Fast, intelligent responses even for complex agricultural questions
• Cost-effective — critical for a free tool targeting low-income farmers
• Reliable AWS infrastructure ensures uptime in regions across Africa
• Multimodal capabilities open the door to future photo-based pest detection

---

🗺️ Roadmap

• Photo upload for visual pest and disease identification
• SMS fallback via Twilio for farmers without smartphones
• Real-time local market price data integration
• Offline mode with cached common answers
• Native Android & iOS apps
• Community farmer-to-farmer knowledge sharing

---

👨‍💻 Author

Built with ❤️ for farmers everywhere.

---

📄 License

MIT — free to use, fork, and build upon.

---

🌾 FarmAssist — Because every farmer deserves expert advice.