Derived Market Data Systems Engineering

Engineering optimizations for a derived market data database in a trading system.

During an industry internship, I contributed to improving a derived market data database used in a trading system.

Derived market data refers to market data produced through relatively simple computations on raw market data. Previously, each strategy team computed this data independently; the computations have now been integrated into the trading system.

My primary contributions included:

  1. Adding replay support for derived market data, allowing the system to simulate market activity using persisted data.
  2. Implementing a replay-configuration script to replace the inconvenient process of writing lengthy JSON configuration files.
  3. Extending the derived market data library, which originally supported only operators written in C++, to support Python operators.
  4. Adding support for condition-based triggering in the derived market data library.