Finda Pro-Grade Data Engine v2.0

Institutional-grade async financial data pipeline with Parquet caching, multi-provider fallback, and WebSocket streaming.

Features

• ⚡ Async Architecture - Non-blocking I/O with asyncio and ccxt.async_support
• 💾 Parquet Caching - High-performance local cache with smart data merging
• 🔄 Smart Fallback - Auto-switch providers on failure with health tracking
• 📊 Pydantic Models - Type-safe Candle, Tick, and MarketMetadata schemas
• 🔌 WebSocket Streaming - Real-time tick and candle subscriptions
• 🌍 Multi-Provider - Dukascopy (Forex), Binance (Crypto), Alpaca (Stocks)

Installation

pip install finda

With Alpaca support:

pip install finda[alpaca]

Quick Start

Async (Recommended)

import asyncio
from finda import Finda

async def main():
    f = Finda()
    df = await f.get_candles("EUR/USD", "1m", "2024-01-01", "2024-01-02")
    print(df)

asyncio.run(main())

FastAPI Server

uvicorn main:app --reload

Endpoints:

• GET /ohlcv - OHLCV candles with caching
• GET /tick - Tick-level Bid/Ask data
• GET /markets - Provider health & available symbols
• GET /cache/stats - Cache hit/miss statistics

Legacy Sync

from finda import fetch_unified_ohclv
o, h, l, c, v, t = fetch_unified_ohclv("EUR/USD", "1m", "2024-01-01", "2024-01-02")

Configuration

Create .env file:

ALPACA_API_KEY=your_key
ALPACA_SECRET_KEY=your_secret
CACHE_DIR=.finda_cache
LOG_LEVEL=INFO

Data Sources

Provider	Asset Class	Data Types
Dukascopy	Forex	OHLCV, Bid/Ask Ticks
Binance	Crypto	OHLCV, Trades
Alpaca	Stocks	OHLCV, Trades

License

MIT