Skip to content

ICE Order Book Building Accelerator Overview


The ICE Order Book Building Accelerator constructs order books from ICE (The Intercontinental Exchange) Realtime L2 market data and persists them in kdb Insights Enterprise. Book data can be streamed directly into custom Realtime streaming analytic pipelines, queried using SQL, or analyzed using enhanced Accelerator APIs via Python or any downstream application using REST.

The entire Accelerator can be configured and launched on an existing kdb IE installation in minutes. Quants, business users, and developers can get instant access to high quality ICE data, paired with the power and performance of kdb Insights Enterprise, all without a lengthy implementation project.


kdb IE + ICE Order Book Accelerator can power use cases such as:

  • Backtesting – Providing faster backtesting runs, which means quicker algo development iteration
  • Post-trade analytics and TCA – Gain fine grained access to L2 book snapshots for execution traces and true cost analyzes. Addresses compliance needs and gives better informed order entry strategies.

The Accelerator delivers a scalable, performant architecture including:

  • The ICE Realtime Feed Handler
  • Pipelines for L1 Trade and Quote data ingest icerealtime1-sp
  • Pipelines for L2 Quote data ingest icerealtime2-sp and orderbook construction orderbook-sp
  • Schemas for L1 data (Trade and Quote), L2 data (L2Quote), and constructed order book data
    • Snapshot formating, with arrays of price levels in a single field OrderBookSnapshots
    • Depth formating, with order book levels on individual rows Depth
  • Query and analytic APIs via the FSI Library
  • An example Insight View for displaying streaming data
  • Example Jupyter notebooks, demonstrating Python and REST API access

ICE Order Book Accelerator Architecture

Further detail