How many concurrent users can be handled by RASA at a time ? Is there a method or settings to modify it ? Does the model size impact the no of concurrent hits at a time?
With the recommended deployment methods it should be able to handle 25 requests/sec (or 250 concurrent active users/sec if they send a message every 10 seconds).
The model size does impact these number, so please run your own tests to get more accurate estimation.
Hope that helps!