We cloned a voice, a face and a way of moving — and nobody noticedAlways in the Room

A face this familiar
leaves no room for error

Cloning a public figure who already has an audience is the hardest version of this problem. Every micro-expression, every breath before a sentence, every way she tilts her head toward a swatch of paint had to survive the ultra-close-up — the exact distance where synthetic video usually breaks.

There was no room for the uncanny valley. Not for seven seconds of it, not for one frame. The brief was binary: either viewers believe it's her, or the personal brand loses its most valuable asset — trust.

• A recognizable face with years of real footage it had to match, frame for frame.
• An accented speech pattern and cadence that had to survive voice cloning untouched.
• Interior-design vocabulary — texture, light, material — that needed to sound like expertise, not a script.
• Enough footage variety to cover portfolio pieces, course intros and product pitches without repeating a single gesture.

So we built her a double
that doesn't perform being her

It just is her, frame by frame — the same artist, freed from the limits of a single camera and a single day.

The approach split into three layers that had to agree with each other: a cloned voice trained on hours of her real speech, a face model rebuilt for ultra-close-up honesty, and a movement layer that keeps her hands, eyes and posture doing what they would do off-camera too.

She is a real interior artist. The film exists to prove her digital double can carry that reality into every room her content reaches — a portfolio walkthrough, a course intro, a client pitch, on repeat, without asking her to sit down and record again.