GOODJUJU

GoodJuju: Video production without the production bottleneck

Two pipelines that turn raw material into finished, publishable video automatically — cuts, sync, narration, assembly. The editorial craft of a 20-year VFX career, encoded into systems that produce at the speed and cost of software.

Industry

Internal system · productized

Client

GoodJuju

Year

2026

50s · THIS AD WAS PRODUCED BY THE PIPELINE IT SHOWS
50s · THE BREACH — A SHORT FILM PRODUCED END-TO-END WITH AI, UNDER REAL DIRECTION

The Challenge

Video is the growth channel, but editing is the bottleneck: a skilled person spending days on mechanical cuts, or an explainer video costing hundreds in production time. The economics capped the output.

Our Approach

The editor pipeline transcribes with Whisper, detects speech with Silero VAD, and cuts silences and bad takes with FFmpeg — 99 minutes of raw footage became a 71.9-minute edit with 1,250 cuts, sync within ±0.021s.

The explainer pipeline runs a 9-step chain — script, narration, animated presenter, motion graphics, assembly — producing a finished video in 30–45 minutes for roughly $14.

Both run as repeatable systems: the same pipelines that produce GoodJuju's own content are the product clients can buy.

Outcomes & Impact

1,250 — cuts in a single automated edit. ±0.021s — audio/video sync precision. ~$14 — in API cost per finished explainer video. 30–45 min — per video, end to end.

In production use on real content.

AI is here. Most will react.The few with a plan will lead.

We build for those few.