I have carla running at all times and put all of my system audio through various loopback devices (browser, voice conferencing, and music/system) and then apply varying degrees of compression to them (no surprise sounds, hard limiting to hear quiet people, and bypass -- respectively).
Of course everything goes through an extra limiter at the end to avoid clipping.
I also send the voice conferencing input and output through RNNoise, so I can avoid emitting terrible sounds and avoid hearing them as well.
People also seem to like me better when I cut my mids a little bit, but additional research is required.
The reason for this is that I can change browsers (or games or voice apps) and they all think I'm just using a normal mic and headset, but it's actually like 10 LSP plugins and various routing.
Still doesn't feel that complicated when you do a little bit at a time.
Instead of digitally recreating your voice so people like you more, have you considered getting some psychotherapy? Maybe? :/