nt-architect

To translate ML models into a NautilusTrader architecture, you map research artifacts to Actors, Indicators, and custom data types, defining component ownership and data contracts before implementation.

Designing data flow in NautilusTrader involves defining how research outputs map to Actors and custom data types, using recommended message bus patterns and planning state initialization for data flow modeling.

You need an architecture document during initial system design for new trading architectures to capture phase definition, data flow patterns, state management, and integration sequence before coding.

Yes, component decomposition takes research outputs and breaks them down into a structured architecture document detailing intake, data contracts, and integration sequence for downstream implementation.

Yes, NautilusTrader architecture design requires state initialization and state management plans to define how data flows through Actors and Indicators within the component architecture.

Skipping architecture design risks unclear component ownership and data contracts, leading to integration failures during implementation due to poorly defined data flow patterns and state management.