Centralized AI Processing for Multiple Remote Nx Witness Recording Servers

Hello Nx Team,
We are currently designing a large-scale video surveillance system using Nx Witness, with multiple remote sites (stations) distributed across different geographical locations and connected to a central monitoring center.
Our proposed architecture:
1. Remote Sites (Stations)
Each remote site has a local server running Nx Witness.
The local servers are primarily responsible for camera management and video recording.
Due to limited hardware resources (CPU/GPU), these servers are not powerful enough to perform AI video analytics.
We would like to keep the remote servers dedicated to video recording without running AI processing locally.
2. Central Monitoring Center:
All remote Nx Witness servers are connected to the central monitoring center via the network.
We would like to merge or centrally manage all remote Nx Witness servers within a single Nx Witness system.
We plan to deploy one or more dedicated AI servers at the central monitoring center.
These central AI servers would receive video streams from cameras at the remote sites and perform AI analytics centrally.
AI detection results and events should be integrated into Nx Witness for centralized monitoring, notifications, and event management.
Our questions:
Does Nx Witness support this type of distributed recording architecture with centralized AI processing?
Thank you for your support.