Dispatch № 02 · crypto-engine
Crypto Trading Engine
Trading engine across multiple exchanges
- Python
- TimescaleDB
- PostgreSQL
- AWS S3
- Parquet
The core components of a high performance, modular engine that connects to several cryptocurrency exchanges at once and provides the building blocks for market data, order management, and backtesting.
By the numbers
- Built the core of a cryptocurrency trading engine in Python that connects to several major exchanges at once, aggregating live order book data and funding rates behind one interface.
- Implemented unified order placement and cancellation, real time position and PnL tracking, a universal symbol mapper, and historical order book persistence for backtesting.
Market data aggregation
Modular connectors pull best bid and ask prices, full order book depth, and both current and historical funding rates from several exchanges, plus price impact analysis for a given trade size to surface arbitrage opportunities.
Order and position management
A single interface places and cancels limit and market orders across exchanges with latency tracking, then monitors the resulting positions and computes real time unrealized PnL against live market data.
Symbol mapping
A universal symbol mapper normalizes the different ticker formats and quirks each exchange uses into one consistent representation, so the rest of the engine never has to special case a venue.
Persistence
A configurable ingestion service captures live order books into TimescaleDB for time series storage and into S3 as partitioned Parquet files, giving a clean historical dataset for backtesting.