VIDME.

IN DEVELOPMENT

Less repetitive editing.
More room for creativity.

Vidme is an AI-powered narrative discovery engine. We leverage LLM transcript analysis to automatically extract high-converting 10-second hooks and edit specs directly from raw video timelines.

Explore the project
01 ABOUT THE PROJECT

Why I'm building Vidme

Manual transcript parsing and hook discovery consume hours of editing time. Vidme is being developed with Google Gemini API integration, combining AI-driven narrative analysis with Silero VAD and WhisperX audio processing to identify promising 10-second hooks and generate structured editing specifications.

TOOLS TECH STACK

Tech Stack

  • Transcript & VAD Parsing: Silero VAD + WhisperX for speech detection and transcript alignment.
  • LLM Hook Discovery: Google Gemini API integration for narrative analysis, pacing evaluation, and retention-focused hook discovery.
  • Timeline Spec Export: Structured JSON outputs for automated editing workflows and NLE integration.
02 WHERE IT STANDS

A work in progress

Vidme is being developed through practical editing experiments and production workflows. The product is not publicly available yet. This website will grow as the project develops.

Need a video editor today?

Explore Level Eight

GET IN TOUCH

Have a project in mind?

For project inquiries or conversations: