Voodoo.tube turned long-form video into captioned short clips. Ego Eimi built the product with Abhishek Gawade, connecting the analysis pipeline to rendering and export. The product was later acquired.
A creator repurposing a long recording needs to find an appropriate moment, cut it, reframe it and add timed captions. Selecting a passage without producing the finished video leaves much of the work untouched.
The product needed a connected pipeline that could handle a source file and deliver usable output, with job processing suited to media workloads.
The application accepted a URL or upload, transcribed the audio and analyzed the content to propose moments for clipping. The rendering pipeline produced vertical clips with styled, timed captions.
Next.js and TypeScript supplied the application, while Python, speech-to-text, LLMs, FFmpeg and job queues supported processing. Clip scores helped users compare suggestions, with export completing the workflow.
Ingestion, analysis, rendering and export. The delivery section explains the workflow demonstrated here.
The project delivered the full path from source video to a finished clip, and the product later changed ownership through an acquisition. That history connects the delivery to the practical requirement for software to remain understandable beyond its original team.
An acquirer needs to understand processing queues, storage, model providers, rendering infrastructure and usage costs. Access to the repository is only one part of that transfer. The operating record should show how a job is recovered and how the buyer can change the pipeline after closing.
Project contribution: Abhishek Gawade. The scope described above identifies the work behind this case.
Editorial review: Ego Eimi · Updated September 6, 2026 · About the team
Your technology
Start with the business
Tell us which system matters, what has changed and who controls it. We will define a reviewable next step. Read Foundation and the commitments and evidence.
