Hi,

I’ve been testing your platform with a medical examination video involving two people: the practitioner and the patient. The issue is that your AI consistently focuses on the practitioner’s face instead of the actual action happening on the patient.

In this type of content, the interesting and engaging part is not the person speaking or standing there, but the hands-on treatment/examination being performed on the patient. That is where viewers naturally focus their attention.

Your clipping system currently seems to prioritize faces and speaking patterns as “viral signals,” but in many real-world videos — especially medical, treatment, reaction, or procedure content — the important visual focus is the interaction itself, not the provider’s face.

Because of this, the generated clips completely miss the most engaging moments of the video.

It would be extremely useful to have:

* a way to manually define the subject/focus of the clip,

* object/action prioritization instead of face prioritization,

* or an option to tell the AI which person should remain the center of attention.

Right now the clips feel disconnected from what actually makes the video interesting.

Thanks.