On-fly speech processing for last few seconds

[lukypko]

I like that application, I haven't try to build it manually to try the latest source code (I'm just using your last released version from 2021-04-05).

Would it be possible to create an on-fly speech processing, for example, showing information like speech rate, articulation rate, phrase pauses for a recent few second, for example for recent 3, 10, 20 seconds.
Then a speaker can compare those values and see: "Aaa, I started to speak too fast, I have to slow down little bit",
then based on values from last 3 seconds and 10 seconds a speaker can compare if his tempo was slow that before.