Video Engine
A faceless-video pipeline sourcing business-failure stories from court filings.
See it run
The Problem
Faceless business/startup-failure content is crowded, and everyone sources from the same recycled articles. The edge is in the sourcing — going to the primary filings nobody else pulls.
What I Built
A pipeline reusing TechBeat's court-filing infrastructure (8-Ks, bankruptcy dockets) as a sourcing moat: Python sourcing feeds a human script edit, then Canva MCP autofills locked episode and thumbnail templates, edge-tts and faster-whisper and ffmpeg handle voiceover, captions, and render, and the YouTube API publishes.
The Outcome
In progress. Sourcing and render architecture are defined; next step is building and locking the Canva templates around the brand kit. Targeting 1–2 quality videos per day.
Like what you see?
Let's build something.