Lensward

Lensward is a rehearsal coach for speaking to camera — a job interview on video, a piece to camera, a pitch you want to land. Prop your phone up, look at the lens, and run it. Lensward reads how you actually came across, and shows you where your delivery slipped.

It uses the TrueDepth camera the way Face ID does: rather than recording video, it reads 52 facial muscle signals, your head pose and where your eyes are pointed, sixty times a second. From that it measures four things an audience registers before they hear a word — how much you held the lens, how much your face moved, your blink rate under nerves, and how still you held your head. On-device speech adds your pace and filler words.

Every measure is scored against a target range, not a ceiling. Unbroken eye contact reads as a stare; a motionless head reads as stiff. Because each measure has a healthy range, the report can flag when you overshot it, not only when you fell short — and it leads with the moments, timestamped, so you know exactly when to fix something, not just that something was off.

• Rehearse — first 30 seconds, elevator pitch, interview answer, piece to camera, or open practice

• A live 52-channel read of your face while you run

• A report that names the moments, then the numbers

• Learn — a library of short technique lessons

• Journal — private, encrypted notes to yourself between takes

• Practice — streaks and a calendar of your sessions

Private by design. There is no account and no network — video and audio are analysed frame by frame and discarded, nothing is recorded, and nothing ever leaves your device. Your reflections are encrypted on the device that wrote them.

Lensward needs an iPhone with a TrueDepth (Face ID) camera.