Algorithmic traders waste hundreds of hours stitching together fragmented exchange APIs, handling WebSocket drops, and normalizing TradFi versus Crypto data structures. SpanTrad abstracts this chaos into a unified, reliable data and execution pipeline.
SpanTrad is built around a secure execution mesh designed to protect your intellectual property, delivering a seamless, engaging and powerful experience as you deploy your algorithms into live markets.
Algorithms can be written using our headless Python SDK (pip install spantrad) and executed by your machine's locally installed Python, or fully compiled into standalone executables before deployment if preferred.
Communication with the core trading engine occurs via encrypted WebSockets, ensuring your proprietary trading logic remains strictly isolated from the execution layer.
Dynamic GUI rendering allows your headless bots to instantly project custom control panels into the desktop application.
Although SpanTrad is designed as an algorithm-centric terminal, it is a powerhouse fully capable of simultaneous discretionary trading.
Extend your functionality as a trader even further by making use of drawing-based alerts, which can be configured to place trades, start or stop bots, launch external scripts, and initiate both private and public Telegram messages.
Build your historical data library from multiple sources. Our intelligent stitching engine allows you to compile your database seamlessly.
Additionally, our local DuckDB integration powers lightning-fast, zero-copy backtesting without network bottlenecks, meaning you can pull data and backtest your ideas inside a notebook (e.g., Jupyter) with ease. You can even choose whether to call 'stitched' or 'unstitched' data for scenarios where you wish to use cleaned, institutional-grade data exclusively in your backtests.
SpanTrad is built for quantitative developers, algorithmic researchers, and systematic traders who require raw execution control, dual-mode risk routing, and absolute data privacy.
SpanTrad is currently in an alpha state. The core engine and SDK are undergoing rapid internal testing, and public access is currently restricted to core contributors.