Provisional Patent — US 64/100,191
Video, cut to the sound of speech.
Synctia is the sync engine behind AI-generated video and audio — narration, subtitles, and music locked together to the millisecond, shipped as an API/SDK for the tools creators already use. Not by waiting for speech recognition to finish a word, but by detecting the instant it begins.
The problem
Every existing tool syncs too late.
How it's done today
Captioning and dubbing tools time text to the moment speech recognition finishes a word — after it has already been spoken. The gap is small, but it's audible and visible: a caption that lands a beat behind the mouth that formed it.
What Synctia does instead
Synctia detects the onset of the consonant — the attack — the same reference point a musician uses to lock a performance to a click track. Subtitles, cuts, and music land on the sound, not after it.
The product
Shipped where creators already work.
Synctia ships as an API/SDK, distributed through creative-tool marketplaces like Adobe Exchange and Canva Apps. Creators get frame-accurate sync inside the tools they already use — no new software to learn, no workflow to change.
Where we are
Status
-
01
Provisional patent filed USPTO No. 64/100,191 — consonant attack point synchronisation27 JUN 2026
-
02
Synctia Ltd incorporated Company No. 17285447, England & Wales19 JUN 2026
-
03
Innovate UK application submitted Next Wave: Breakthrough — Wave 1, awaiting decision03 JUL 2026
-
04
SEIS Advance Assurance submitted HMRC review in progress03 JUL 2026
Founder
Built by someone who heard the problem first.
Toshitaka Fujikawa
A composer and musician before a founder. Years of scoring and editing video by ear — nudging captions and cuts back onto the beat one frame at a time — led to the idea underlying Synctia: sync video the way a musician syncs a performance, off the attack, not the aftermath.