Hi,

Assume that a machine learning model is fitted to past data and saved in object-store. Then a live trading algo is launched which uses the fitted ML model.

When live algo is currently running, is it possible to use research environment and refit ML model, save it to object-store, so that the live algo can access to it?

It all depends on whether live trading has access to objects in object-store that are written after the deployment or not. 

Thanks