Persistent character identity
One canonical identity per character, reused as a reference in every downstream generation.
Building the story-to-video engine
Long-form video generation is not blocked by image quality any more. It is blocked by memory. Voiceido builds the representation that lets a whole book render as one coherent visual story.
01
Problem
A book contains hundreds of interconnected story facts: who exists, who is present, what they look like, where they are, what already happened, and what is only being remembered.
A prompt-by-prompt video tool has no memory. Ask it for the same character in scene 1 and scene 28 and you get two different people — different face, different clothes, different age. That is tolerable for a single clip and fatal for a story.
Every additional minute of video multiplies the number of chances to break continuity. That is why generative video demos are short, and why adaptations are still made by studios.
02
Approach
Extract the story first. Generate last.
- 01BookEPUB, PDF, DOCX, or manuscript text
- 02Story understandingPlot, chronology, and relationships
- 03Character registryOne canonical identity per character
- 04World registryLocations and props that persist
- 05Scene graphShots planned against the source text
- 06Visual generationFrames locked to the cast references
- 07MotionShot-level animation
- 08NarrationVoice measured first, then timed to picture
- 09VideoCaptioned, assembled, downloadable
03
Differentiation
Story-level understanding
The whole book is read and structured before a single frame is planned.
Location and prop continuity
Places and objects are registry entries, not adjectives retyped per shot.
Provider independence
Models sit behind one adapter, so the system improves as the ecosystem improves.
Human approval workflow
Cast and scenes are approved before paid generation, which is what makes the output usable commercially.
Long-form orchestration
Durable, resumable, concurrent production across an entire book, with per-asset cost accounting.
04
Who this serves
Ordered by how quickly they can adopt it, not by market size. We do not publish market-size estimates we cannot source.
Authors
Self-published and independent authors who own their rights and need visual marketing.
Publishers
Backlist and series promotion at a cost per title that scales.
Education
Public-domain and licensed texts turned into narrated visual material.
Audio and podcast creators
Existing narrated catalogs that need a visual edition for video platforms.
Creator economy
Story channels that publish on a schedule and cannot afford per-episode animation.
05
Where the product actually is today
Stated plainly, because inflating this is the fastest way to lose a technical diligence conversation.
- Full books run end to end in production: ingestion, story understanding, cast, scene planning, narration, motion, and packaged video.
- Production runs on a durable cloud pipeline with leases, checkpoints, idempotent paid operations, and per-asset cost recording.
- Multiple production packs run concurrently against pooled provider capacity.
- Author pilots are the current focus. Traction metrics will be published here only once they are real and measured.
Contact
Technical diligence, product questions, and introductions reach the founder directly at aliyev@mit.edu.
Keep reading
About VoiceidoVoiceido is building an AI-native production system that turns long-form written stories into consistent visual narratives.How Voiceido WorksThe technical architecture behind Voiceido: story understanding, persistent character and world registries, scene graphs, and voice-timed rendering.Maintaining Character Identity Across AI-Generated StoriesWhy generative video loses characters between scenes, and how Voiceido keeps one canonical identity per character across an entire book.See Voiceido in Sixty SecondsA book becomes story understanding, a locked cast, planned scenes, motion, and narrated video. See each stage without creating an account.