Show HN: I built a camera that you talk with instead of seeing your own face

I used to hate recording myself, but at the same time I also wanted to post more video i.e. on X.

My usual experience was: Get an idea -> reluctantly hit record -> become self-aware, stumble, do 8 takes -> give up.

So I tried changing the camera instead.

MakiRoll doesn't show you your face while you're recording.

You give it a rough topic and it asks you questions, or lets you ramble.

Those questions adapt to what you've just said via STT and a language model, so recording feels more like having a conversation than performing at a lens.

Once you're done, it looks through the transcript, finds the interesting takes, suggests a structure and makes a captioned cut.

The interesting part for me hasn't really been the editing part, which is nice, but by removing the mirror and replacing it with another participant seems to change how people behave while recording.

One of my early testers described himself as super shy, and now he's making videos and getting more comfortable doing it.

It's built for web + iOS. I'd love some feedback on whether the conversational recording model makes sense to people who normally avoid going on camera (but would like to do that more).