Rowy integration delay?

I was starting to work around with the OpenAI Assistant (Stream Response) node. This AI shall use personality parameters that are saved externally (in my case on Directus). Now I might not want to fetch a request every time a message is sent to that endpoint, so I was wondering if I should rather save these parameters in rowy and use Buildship's integrated Nodes. Though I am not sure if it would have the same latency compared to my Directus GET request. Does BuildShip internally also use REST API calls to fetch data from Rowy? I never used Rowy before and do not really understand the code behind the such nodes