- React
- Framer Motion
- Product Design
- Narrative Design
Narrative Intelligence System
Continuity Atlas
Narrative Intelligence for Complex Fiction
Most writing software stores documents. Continuity Atlas stores relationships — it turns a manuscript into a searchable narrative knowledge graph.
The problem
Long-form fiction breaks in surprisingly mundane ways.
Characters change eye color between chapters. Secondary characters disappear for fifty pages before returning without explanation. Timelines drift. Languages evolve inconsistently. Religious systems accumulate contradictions. Entire plotlines become victims of revision.
Traditional notes quickly become unsearchable, while generic AI tools lack awareness of the living internal logic of a manuscript. For projects spanning hundreds of pages — or multiple books — the cost of continuity errors grows exponentially.
Who it’s for
Designed for
- Novelists writing large-cast fiction
- Literary horror authors
- Multi-book series authors
- Developmental editors
- Publishing teams maintaining franchise consistency
- Worldbuilders creating internally coherent universes
A typical project
- 75–300+ named characters
- Multiple timelines
- Invented languages
- Religious systems
- Genealogies
- Geographic histories
- Repeated symbolic motifs
- Cross-volume continuity
The solution
Instead of static notes, the platform models the relationships between every major narrative element. Every revision updates the narrative model rather than creating disconnected notes.
- Character profiles
- Family trees
- Timeline events
- Object histories
- Location references
- Linguistic evolution
- Religious traditions
- Symbol recurrence
- POV tracking
- Chapter appearances
- Dialogue attribution
- Plot dependency chains
Interactive prototype
A working prototype built on real manuscript data from Liminal 6:17 — inspect Story Memory, Character Drift, the Voice Fingerprint, and the Rewrite flow.
Core features
Character Intelligence
Every character becomes a living record rather than a paragraph of notes.
- First appearance
- Relationships
- Aliases
- Physical descriptions
- Dialogue patterns
- Injuries & deaths
- Knowledge boundaries
- Unresolved arcs
Timeline Engine
Builds an internally consistent chronology and flags what can't be true.
- Impossible travel
- Overlapping events
- Age inconsistencies
- Seasonal conflicts
- Duplicated scenes
Continuity Validation
Automatically detects contradictions across the manuscript.
- Changed descriptions
- Conflicting dates
- Impossible knowledge
- Repeated introductions
- Forgotten subplots
- Inconsistent terminology
World Bible
A continuously updated reference that rebuilds itself with every revision.
- Locations
- Organizations
- Religions
- Fictional languages
- Artifacts
- Historical events
- Laws & customs
Semantic Search
Instead of searching for words, authors search for concepts.
- every conversation involving Rowan before Chapter 8
- references to Brynjavík churches
- all mentions of Vikamál grammar
- scenes involving mourning rituals
Technology
Frontend
- React
- TypeScript
- Tailwind CSS
Backend
- Azure Functions
- Node.js
- GraphQL APIs
Storage
- Azure SQL
- Azure Blob Storage
AI
- OpenAI APIs
- Vector embeddings
- Semantic retrieval
- RAG architecture
Supporting
- Azure Cognitive Search
- GitHub Actions
- CI/CD pipelines
Outcome
Authors spend dramatically less time hunting through previous chapters. Editors review a continuously updated continuity model instead of static notes. Worldbuilding stays internally consistent even across extensive revisions.
Revision cycles become substantially faster because continuity verification is automated alongside traditional editing.
Example use cases
- Literary HorrorMaintain recurring symbolic systems across an entire novel.
- Epic FantasyTrack hundreds of named characters over multiple generations.
- Historical FictionValidate chronology against real historical timelines.
- MysteryTrack clue placement, reveals, suspects, and information exposure.
- Multi-Series PublishingMaintain continuity across novels sharing one universe.
Why it’s different
Rather than treating a manuscript as pages of text, it models the narrative itself — characters, events, places, motifs, languages, and dependencies — as interconnected data that can be queried, validated, and explored.
Less time searching through notes. Fewer continuity errors. A more reliable foundation for developing complex fictional worlds.