# SceneVerso Full Specification & Architecture
Website: https://sceneverso.com
Organization: Golden Corpus (Tokyo, Japan)

## 1. Product Summary
SceneVerso is a quality-gated video production system built for consultants, founders, and technical professionals who demand high-fidelity video without credit-burn waste.

## 2. The 5-Stage Architecture
1. Ingestion & Voice DNA: Reads user writing profiles (MyWritingTwin RuntimeBlock v2) and brand color/typography definitions.
2. Manifest Generation: Deconstructs the narrative into a discrete scene manifest (duration, visuals, voice narration, audio beds, transition effects).
3. Human-in-the-Loop Approval Gate: Renders a zero-cost storyboard with estimated character count and duration. The user explicitly approves the manifest before any provider API is touched.
4. Production: Dispatches narration to ElevenLabs / Whisper and visual prompts to specialized models.
5. Remotion Compositing & Quality Gate: Combines visual tracks, voiceover, music beds, and subtitles with automated WPM and motion continuity auditing.

## 3. Unit Economics & Pricing Model
In Phase 0 dogfood benchmarks across 5 real production videos, measured cost per finished minute was $0.24 (TTS + music).
Traditional platforms charge $20-$100/mo for credit pools that vanish on failed generations.
SceneVerso positions with flat-rate packaging ($29-$49 one-time CLI license) with Bring-Your-Own-Keys (BYO keys) for underlying generative providers.

## 4. Siblings in the Golden Corpus Ecosystem
- MyWritingTwin (mywritingtwin.com) - AI Writing Profiles & Voice DNA
- FluxDiagram (fluxdiagram.com) - Animated diagrams for presentation decks
- SlideVerso (slideverso.com) - Presentation slides generation