Stock Intelligence Copilot

Overview

An AI-assisted stock market analysis system for retail investors. Provides probabilistic insights, never guarantees. Read-only by design.

Core Principles

• Assistive, not automated: Suggestions only, no trade execution
• Probabilistic: Confidence scores, never certainty
• Explainable: Clear reasoning for every insight
• Risk-aware: Built-in safety constraints

Architecture

backend/
├── app/
│   ├── api/              # FastAPI routes
│   ├── core/             # Core business logic
│   │   ├── market_data/  # Data ingestion
│   │   ├── indicators/   # Technical indicators
│   │   ├── signals/      # Signal generation
│   │   ├── risk/         # Risk assessment
│   │   ├── explanation/  # Reasoning layer
│   │   └── orchestrator/ # Pipeline coordinator
│   ├── models/           # Pydantic data models
│   └── config/           # Configuration
├── tests/                # Unit tests
└── main.py               # Application entry point

MVP Capabilities

• Analyze stock data (price, volume, indicators)
• Generate Bullish/Bearish/Neutral signals with confidence
• Explain reasoning and assumptions
• Enforce risk constraints
• Long-term investing focus (default)

Not Included in MVP

• Trade execution
• Exact price predictions
• Real broker API integration
• Frontend UI
• User authentication

Getting Started

# Install dependencies
pip install -r requirements.txt

# Run the server
python main.py

API Endpoints

• GET /health - Health check
• POST /api/v1/analyze - Analyze a stock
• GET /api/v1/supported-tickers - Get supported tickers

Safety Notes

• All outputs are suggestions, not financial advice
• No guarantees of returns or accuracy
• "No action recommended" is a valid outcome
• Users must do their own research