🌾 Agentic AI Agriculture Chatbot

A modern, intelligent agriculture chatbot powered by LlamaIndex and Ollama that provides farmers and agriculture enthusiasts with comprehensive information about farming practices, weather updates, and market prices.

🚀 Features

• Agriculture Knowledge Base: Vector-based retrieval from agricultural documents and farmer handbooks
• Real-time Weather Updates: Get current weather conditions for any location using OpenWeatherMap API
• Market Data Integration: Access live mandi prices from e-NAM (National Agriculture Market) platform
• Intelligent Agent: ReAct agent that can reason about which tools to use based on user queries
• User-friendly Interface: Streamlit-based web interface for easy interaction
• Multi-modal Responses: Combines document retrieval, summarization, and real-time data

🏗️ Architecture

The chatbot uses a modern agentic AI architecture with:

• LlamaIndex: For document indexing and retrieval
• Ollama: Local LLM (Qwen 3:1.7b) for reasoning and response generation
• ChromaDB: Vector database for storing document embeddings
• Streamlit: Web-based user interface
• spaCy: Natural language processing for entity extraction
• HuggingFace Embeddings: Sentence transformers for semantic search

📦 Installation

Prerequisites

1. Python 3.8+
2. Ollama: Install and run Ollama locally

   # Install Ollama from https://ollama.ai/
   # Pull the required model
   ollama pull qwen3:1.7b

3. spaCy English Model:

   python -m spacy download en_core_web_md

Setup

1. Clone the repository:

   git clone https://github.com/Horizon733/Agentic-AI-demo.git
   cd Agentic-AI-demo

2. Install dependencies:

   pip install -r requirements.txt

3. Configure API Keys:

   • Get your OpenWeatherMap API key from openweathermap.org
   • Update OPENWEATHER_API_KEY in utils.py:

   OPENWEATHER_API_KEY = "your-actual-api-key-here"

4. Prepare Documents (Optional):

   • Place your agriculture-related PDF documents in the docs/ folder
   • The system comes with a sample farmerbook.pdf

5. Initialize Vector Database:

   python populate_db.py

🎯 Usage

Running the Chatbot

1. Start Ollama (if not running):

   ollama serve

2. Launch the Streamlit app:

   streamlit run agent.py

3. Open your browser and navigate to http://localhost:8501

Example Queries

• Weather: "What's the weather in Mumbai?"
• Agriculture: "What fertilizer should I use for rice cultivation?"
• Market Data: "Show me today's mandi prices"
• Combined: "What's the weather in Delhi and what crops can I grow there?"

📁 Project Structure

├── agent.py                 # Main Streamlit application with ReAct agent
├── utils.py                 # Utility functions for weather, market data, and indexing
├── populate_db_doc.py      # Script to populate vector database from documents
├── requirements.txt        # Python dependencies
├── docs/                   # Agriculture documents folder
│   └── farmerbook.pdf     # Sample agriculture handbook
├── chroma/                # ChromaDB vector store (auto-generated)
└── storage/              # LlamaIndex storage (auto-generated)

🔧 Configuration

Model Configuration

The chatbot uses Qwen 3:1.7b by default. You can change the model in agent.py:

llm = Ollama(model="qwen3:1.7b", request_timeout=60)

Embedding Model

Uses HuggingFace sentence-transformers for embeddings:

embed_model = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")

Vector Store Configuration

ChromaDB settings can be modified in utils.py:

CHROMA_DB_DIR = "chroma"

🛠️ Available Tools

The agent has access to four main tools:

1. WeatherTool: Provides real-time weather information for any location
2. ENAMTool: Fetches current market prices from e-NAM platform
3. docs_vector_query: Searches agriculture knowledge base using vector similarity
4. docs_summary: Summarizes agriculture-related information when needed

🔄 Data Sources

• Weather Data: OpenWeatherMap API
• Market Data: e-NAM (enam.gov.in) trade details API
• Agriculture Knowledge: PDF documents in the docs/ folder
• Web Sources: Can be populated using populate_db.py for web scraping

🚦 API Endpoints

OpenWeatherMap API

• Current Weather: https://api.openweathermap.org/data/2.5/weather
• Geocoding: http://api.openweathermap.org/geo/1.0/direct

e-NAM API

• Trade Details: https://enam.gov.in/capoaservices/tradedetails

🤝 Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add some amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

📝 License

This project is open source and available under the MIT License.

🙏 Acknowledgments

• LlamaIndex for the powerful RAG framework
• Ollama for local LLM inference
• OpenWeatherMap for weather data
• e-NAM for agriculture market data
• HuggingFace for embedding models

📞 Support

If you encounter any issues or have questions:

1. Check the Issues page
2. Create a new issue with detailed information
3. Join our community discussions

🔮 Future Enhancements

• Add more agriculture data sources
• Implement crop recommendation system
• Add multilingual support
• Integration with IoT sensors for real-time farm data
• Mobile app development
• Voice interface support

---

Made with ❤️ for the farming community