BigData side-channel?

We have a compute framework which is designed to process TBs of data in real-time. This compute framework uses a DAG, so it fits into the flow of Node-Red. Now, I do not intend Node-Red to process TBs of data in real time using it's engine, but rather handle the setup of the flow, control and notifications.

Are there examples of creating such nodes, where Node-Red is used to set up the flow paths/connections, but where the actual data processing is happening independently?

And example could be something like:
[input 4k high-speed camera stream] -> [process filter A] -> [process filter B] -> [storage]
Where Node-Red controls the flow connections, and can change the inputs based on events/triggers, but have no hands in the actual dataflow?

Hopefully this makes sense. :slight_smile: