naik123/ai-financial-modeler

AI Financial Modeling Platform

An AI-powered platform for generating institutional-grade Excel financial models for Indian stocks. Built with FastAPI, Next.js, and Google Gemini AI.

[image]


🚀 Features

Core Capabilities

• —AI-Powered Industry Classification: Uses Google Gemini AI to classify companies and select appropriate model templates
• —Real Excel Formulas: Generated models use linked formulas, not static values - fully editable
• —150+ Indian Stocks: Comprehensive database covering Power, Banking, IT, Pharma, FMCG, Auto, and more
• —Raw Data Input: Generate models from your own financial data without scraping

Advanced Financial Modeling

• —Three-Statement Model: Income Statement, Balance Sheet, Cash Flow - all linked
• —DCF Valuation: WACC calculation, FCFF projections, terminal value, implied share price
• —Sensitivity Analysis: WACC vs Terminal Growth, Revenue vs EBITDA Margin tables
• —Scenario Analysis: Bear/Base/Bull case projections with IRR calculation
• —Dashboard with Charts: Revenue & EBITDA bars, margin trends, net income charts

Industry-Specific Templates

• —Power Sector: PLF tracking, fuel costs, tariff analysis
• —Banking: NIM analysis, credit costs, CASA ratios
• —IT Services: Utilization, billing rates, offshore mix
• —Pharmaceuticals: R&D intensity, ANDA filings
• —FMCG: Distribution metrics, brand spending
• —And more!

📋 Requirements

• —Python 3.9+
• —Node.js 18+
• —Google Gemini API Key (optional, uses templates as fallback)

🛠️ Installation

Backend Setup

cd backend

# Create virtual environment
python -m venv venv
venv\Scripts\activate  # Windows
# source venv/bin/activate  # Linux/Mac

# Install dependencies
pip install -r requirements.txt

# Create .env file
copy .env.example .env
# Edit .env and add your GEMINI_API_KEY

Frontend Setup

cd frontend

# Install dependencies
npm install

🚀 Running the Application

Start Backend

cd backend
python main.py
# Server starts at http://127.0.0.1:8000

Start Frontend

cd frontend
npm run dev
# App opens at http://localhost:3000

📡 API Endpoints

Company Information

GET /api/company/{symbol}?exchange=NSE

Returns company name, sector, market cap, and current price.

Stock Database

GET /api/stocks              # All 172 stocks
GET /api/stocks?sector=power # Filter by sector
GET /api/stocks/search/TCS   # Search by symbol or name
GET /api/sectors             # List all 14 sectors

Model Generation

POST /api/model/generate
{
    "symbol": "ADANIPOWER",
    "exchange": "NSE",
    "forecast_years": 5,
    "model_types": ["three_statement", "dcf"]
}

Raw Data Model Generation

POST /api/model/generate-raw
{
    "company_name": "My Company Ltd",
    "industry": "general",
    "forecast_years": 5,
    "historical_data": {
        "income_statement": {
            "revenue": 10000,
            "ebitda": 2500,
            "net_income": 1500
        },
        "balance_sheet": {
            "total_assets": 20000,
            "total_liabilities": 8000
        }
    },
    "assumptions": {
        "revenue_growth": 0.10,
        "ebitda_margin": 0.25,
        "tax_rate": 0.25
    }
}

Job Status

GET /api/job/{job_id}

Download Model

GET /api/download/{job_id}

📊 Excel Model Structure

Each generated model contains:

🏭 Supported Sectors (14 Categories)

1. 1.Power & Utilities - ADANIPOWER, NTPC, TATAPOWER, POWERGRID...
2. 2.Banking - HDFCBANK, ICICIBANK, SBIN, KOTAKBANK...
3. 3.IT Services - TCS, INFY, WIPRO, HCLTECH, TECHM...
4. 4.Pharmaceuticals - SUNPHARMA, DRREDDY, CIPLA, LUPIN...
5. 5.FMCG - HINDUNILVR, ITC, NESTLEIND, BRITANNIA...
6. 6.Automobiles - MARUTI, TATAMOTORS, M&M, HEROMOTOCO...
7. 7.Metals & Mining - TATASTEEL, HINDALCO, JSWSTEEL, VEDL...
8. 8.Oil & Gas - RELIANCE, ONGC, IOC, BPCL...
9. 9.Cement - ULTRACEMCO, SHREECEM, ACC, AMBUJACEM...
10. 10.Infrastructure - L&T, ADANIENT, ADANIPORTS, GMRINFRA...
11. 11.NBFC - BAJFINANCE, BAJAJFINSV, SBICARD, MUTHOOTFIN...
12. 12.Telecom - BHARTIARTL, IDEA...
13. 13.Chemicals - PIDILITIND, UPL, CLEAN...
14. 14.Consumer Durables - TITAN, HAVELLS, VOLTAS, CROMPTON...

🎨 Frontend Features

• —Modern Dark Theme: Glassmorphism effects, gradient accents
• —Stock Browser: Search, filter by sector, click to select
• —Raw Data Input: Enter your own financial data and assumptions
• —Real-time Progress: Live status updates during generation
• —Model Preview: See included features before downloading

🔧 Configuration

Environment Variables

# Required for AI features (optional - falls back to templates)
GEMINI_API_KEY=your_gemini_api_key

# Optional
ALPHA_VANTAGE_API_KEY=your_alpha_vantage_key

Model Assumptions (Editable in Excel)

The Assumptions sheet contains named ranges for:

• —Revenue Growth Rate
• —EBITDA Margin
• —Depreciation % of Revenue
• —Tax Rate
• —Working Capital Days
• —CapEx % of Revenue
• —WACC, Risk-Free Rate, Equity Risk Premium
• —Terminal Growth Rate
• —And more...

📁 Project Structure

ai-financial-modeler/
├── backend/
│   ├── main.py              # FastAPI server
│   ├── requirements.txt     # Python dependencies
│   ├── data/
│   │   ├── yahoo_finance.py # Yahoo Finance data collector
│   │   ├── screener_scraper.py # Screener.in scraper
│   │   └── stock_database.py   # 172 Indian stocks database
│   ├── agents/
│   │   ├── industry_classifier.py # AI industry classification
│   │   ├── financial_modeler.py   # Model structure design
│   │   └── qa_validator.py        # Model validation
│   ├── excel/
│   │   └── generator.py     # Excel model generator
│   └── output/              # Generated Excel files
└── frontend/
    ├── app/
    │   ├── page.tsx         # Main page with stock selection
    │   ├── layout.tsx       # Root layout
    │   └── globals.css      # Dark theme styles
    ├── package.json
    └── tailwind.config.js   # Custom theme config

🤝 Contributing

Pull requests are welcome! Please ensure:

1. 1.Code follows existing style patterns
2. 2.New features include appropriate tests
3. 3.Documentation is updated

📄 License

MIT License - see LICENSE file for details.

🙏 Acknowledgments

• —Yahoo Finance for financial data
• —Screener.in for detailed Indian stock data
• —Google Gemini AI for intelligent classification
• —openpyxl for Excel generation