All Work

Vid Ignite

Script-aware video overlays ~ AI image models that place the right visual on the right word, saving editors hours on every cut.

Services

Web App + API Development

Platforms

Cross Platform

Date

Feb 2025

Status

Retired, now part of a larger project

Image GPT Web Cloud Scalability

Challenge

Generate accurate video overlays from a script at scale, across styles, frequencies and languages.

Traditionally every overlay is found by hand: images sourced, backgrounds removed, each one placed on the timeline and its sound effect timed. On a single video that can take hours.

Solution

With OpenAI's image generation API and Whisper, editors simply drag and drop a video and Vid Ignite returns a precise, word-level edit with every overlay in place.

Whisper transcribes and times each element to the exact word, ElevenLabs powers the sound effects, and a scalable Python pipeline stitches everything into a ProRes file that drops straight into any editor. The result: hours of editing saved on every video.