How can I keep my avatar consistent across AI-generated images and videos?

I'm trying to maintain a consistent avatar across my content, keeping the same face and overall appearance in every image and video. Does anyone have any tips or recommended workflows for achieving this?

My initial plan was to use image-to-image and image-to-video, but so far, image-to-image hasn't been giving me consistent results—or maybe I just haven't figured out how to use it properly yet.

I'd really appreciate any advice or suggestions. Thanks in advance!