Finance Agent 📈

An AI-powered investment analyst agent specialized in Indian stock market analysis. Built with the Agno framework, this intelligent assistant provides comprehensive stock research, fundamental analysis, and automated trading capabilities through Zerodha Kite integration.

🌟 Features

• Stock Market Analysis: Real-time stock prices, technical indicators, and financial ratios using YFinance
• Fundamental Analysis: Deep dive into company financials, income statements, and analyst recommendations
• AI-Powered Insights: Leverages Google's Gemini Flash model with extended reasoning capabilities
• Knowledge Base: RAG (Retrieval Augmented Generation) system with investment lessons from MIT courses and Warren Buffett
• Trading Integration: Place orders on Zerodha Kite platform via MCP (Model Context Protocol)
• Conversational Interface: Natural language interaction with persistent memory
• Indian Market Focus: Specialized for NSE/BSE stocks with proper ticker format handling

🏗️ Architecture

┌─────────────────────────────────────────────────────────┐
│                   Finance Agent                          │
├─────────────────────────────────────────────────────────┤
│  FastAPI Web Application (main.py)                      │
│  ├── Gemini Flash AI Model (8,285 token thinking)       │
│  ├── Reasoning Tools (Analysis & Critical Thinking)     │
│  ├── YFinance Tools (Market Data & Indicators)          │
│  ├── Web Tools (Research & Information Gathering)       │
│  ├── MCP Tools (Zerodha Kite Trading Integration)       │
│  └── Knowledge System (Vector DB with Investment Lessons)│
├─────────────────────────────────────────────────────────┤
│  Data Layer                                              │
│  ├── PostgreSQL + PgVector (Knowledge Embeddings)       │
│  ├── SQLite (Conversation History)                      │
│  └── File Cache (YFinance API Responses)                │
└─────────────────────────────────────────────────────────┘

📋 Prerequisites

• Python 3.8 or higher
• PostgreSQL with PgVector extension
• Google AI API key (Get it here)
• (Optional) Zerodha Kite account for trading features

🚀 Installation

1. Clone the repository

   git clone https://github.com/kamaravichow/finance-agent.git
   cd finance-agent

2. Install dependencies

   pip install -r requirements.txt

3. Set up PostgreSQL with PgVector

   # Install PostgreSQL (if not already installed)
   # For Ubuntu/Debian:
   sudo apt-get install postgresql postgresql-contrib
   
   # Install PgVector extension
   # Follow instructions at: https://github.com/pgvector/pgvector

4. Configure environment variables

   cp .env.example .env
   # Edit .env and add your credentials:
   # - GOOGLE_API_KEY: Your Google AI API key
   # - DATABASE_URL: PostgreSQL connection string

5. Sync knowledge base (One-time setup)

   python sync.py

   This will process all markdown files in the knowledge/ directory and store their embeddings in PostgreSQL.

🎯 Usage

Starting the Agent

python main.py

The FastAPI application will start on http://localhost:8000 with hot-reload enabled.

Interacting with the Agent

Access the web interface or API endpoints to interact with your personal investment analyst:

Example Queries:

• "What is the current price of TCS (Tata Consultancy Services)?"
• "Perform a fundamental analysis of Reliance Industries (RELIANCE.NS)"
• "Show me the analyst recommendations for Infosys"
• "What are the key financial ratios for HDFC Bank?"
• "Place a buy order for 10 shares of TCS on NSE"

Indian Stock Ticker Format

The agent is configured for Indian markets (NSE/BSE). Use the following ticker formats:

• NSE: SYMBOL.NS (e.g., TCS.NS, RELIANCE.NS, INFY.NS)
• BSE: SYMBOL.BO (e.g., TCS.BO)

📁 Project Structure

finance-agent/
├── main.py                 # Main application entry point
├── sync.py                 # Knowledge base synchronization script
├── requirements.txt        # Python dependencies
├── .env.example           # Environment variables template
├── .gitignore             # Git ignore rules
├── knowledge/             # Investment lessons and strategies (markdown)
│   ├── 1.md
│   ├── mit1.md
│   ├── mit2.md
│   ├── mit3.md
│   └── warren-buffet-lessons.md
└── yfinance_cache/        # Cached API responses (git-ignored)

🔧 Configuration

Environment Variables

Variable	Description	Required
GOOGLE_API_KEY	Google AI API key for Gemini model	Yes
DATABASE_URL	PostgreSQL connection string with PgVector	Yes

Agent Instructions

The agent is pre-configured with specialized instructions for Indian stock market analysis. You can modify these in main.py under the instructions parameter.

🛡️ Security Notes

• Never commit .env files to version control
• Database credentials are loaded from environment variables
• The .gitignore is configured to exclude sensitive files
• Review the .env.example for required configuration

📚 Knowledge Base

The agent comes with pre-loaded investment knowledge from:

• MIT OpenCourseWare on quantitative trading
• Warren Buffett investment lessons and principles
• Technical analysis fundamentals

To add more knowledge:

1. Place markdown files in the knowledge/ directory
2. Run python sync.py to update the vector database

🤝 Contributing

Contributions are welcome! Please see CONTRIBUTING.md for details on:

• Code style guidelines
• Development workflow
• Submitting pull requests
• Reporting issues

📝 License

This project is provided as-is for educational and research purposes. Please ensure compliance with:

• Google AI API Terms of Service
• YFinance data usage policies
• Zerodha Kite API terms (if using trading features)
• Securities and exchange regulations in your jurisdiction

⚠️ Disclaimer

This agent is for educational and research purposes only. It does NOT provide financial advice.

• All investment decisions should be made after consulting with qualified financial advisors
• Past performance does not guarantee future results
• Trading stocks involves risk of loss
• The creators and contributors are not responsible for any financial losses

🔗 Resources

• Agno Framework Documentation
• Google AI Studio
• YFinance Documentation
• PgVector GitHub
• Zerodha Kite API

💬 Support

For issues, questions, or contributions:

• Open an issue on GitHub Issues
• Check existing documentation and README first
• Provide detailed information when reporting bugs

---

Built with ❤️ using Agno and Google Gemini