A Bloomberg Terminal-style quantitative finance platform. Full-stack — React/TypeScript frontend with a dark terminal aesthetic, Flask REST API backend with a 350k+ instrument SQLite database, and 11 live quant analysis modules.

Bloomberg chessboard layout: full-width live chart, scrolling world indices ticker tape, live news feed, watchlist panel, company info (4 tabs: Overview / Supply Chain / Analyst / News), sector heatmap, order flow, correlation matrix, earnings history, economic calendar, and 6 draggable floating panels.

QuantLab/
├── backend/          Flask API + SQLite DB
│   ├── app/
│   │   ├── database/     SQLAlchemy ORM (11 tables, 350k+ instruments)
│   │   ├── routes/       14 API blueprints
│   │   ├── services/     Business logic + yfinance data layer
│   │   └── utils/        Response helpers, validators
│   ├── setup_db.py       One-time DB seed script
│   ├── wsgi.py
│   └── README.md         Full backend API reference ← start here
│
└── frontend/         React + TypeScript + Vite + Tailwind
    └── src/
        ├── api/          Typed API client (client.ts)
        ├── components/
        │   ├── layout/   TerminalHome, WatchlistManager
        │   ├── common/   Modal, Panel, DraggablePanel
        │   └── modules/  11 analysis module components
        └── index.css     Bloomberg dark theme + ticker-tape animation

cd backend
python -m venv .venv
source .venv/bin/activate          # Windows: .venv\\Scripts\\activate
pip install -r requirements.txt

# Seed the instrument database (run once — downloads 350k tickers)
python setup_db.py
# Or for instant start with 250 built-in tickers:
python setup_db.py --quick

python wsgi.py
# → http://localhost:3000

cd frontend
npm install
npm run dev
# → http://localhost:5173

• Flask 3.0 — REST API with CORS
• SQLAlchemy — ORM with SQLite (WAL mode for concurrent reads). PostgreSQL-ready via DATABASE_URL env var
• FinanceDatabase — 350k+ global instruments across Equity, ETF, Crypto, Fund, Index asset classes
• yfinance 1.2+ — Live market data. Uses fast_info (chart endpoint) as fallback when quoteSummary returns 401
• PyPortfolioOpt — Portfolio optimisation
• statsmodels — Cointegration tests for pairs trading
• scikit-learn — ML models for price prediction

• React 18 + TypeScript — Component architecture
• Vite — Dev server on port 5173
• Tailwind CSS — Utility-first styling
• TanStack Query — Server state management with automatic refetch intervals
• lightweight-charts — TradingView-grade candlestick charts
• Lucide React — Icons

Prices and market cap figures are displayed in the instrument's native currency. Supported: USD, EUR, GBP, JPY, INR, CNY, HKD, CAD, AUD, CHF, KRW, BRL, and 20+ others. Currency is sourced from yfinance's financialCurrency or fast_info.currency field.

All market data is cached in-memory on the backend with TTLs:

A CLEAR CACHE button in the bottom-right corner of the terminal wipes all cached data and forces a fresh fetch from Yahoo Finance. Individual tickers can also be refreshed via POST /api/terminal/company/<ticker>/refresh.

python setup_db.py             # Full — 350k+ tickers from FinanceDatabase
python setup_db.py --quick     # Fast — 250 built-in tickers (no internet needed)
python setup_db.py --force     # Wipe instruments and re-seed
python setup_db.py --status    # Show stats without seeding

The seeder uses INSERT OR IGNORE to handle duplicate (ticker, exchange) pairs across asset class datasets. Each asset class is seeded in its own SQLAlchemy session so a failure in one doesn't poison the others.

See backend/README.md for the full endpoint reference.