
Novel OS is a voice-first story writing and world-building environment built for novelists who want to stay in creative flow without losing track of the details that make a story work. Writers dictate scenes directly into the microphone. AssemblyAI transcribes the audio in real time, turning raw speech into clean prose. From there, a tiered AI pipeline goes to work, automatically extracting characters, locations, events, relationships, plot threads, open questions, and secrets from the transcript. Nothing is stored as flat notes. Every piece of information is committed to a structured story world knowledge graph, where each fact is tagged with its confidence level and traced back to the exact dictation session it came from. The result is a living story bible that builds itself as you write. You can browse your character roster, explore how relationships evolve across scenes, trace a plot thread through the timeline, or pull up every fact the story has established about a location, all without ever leaving the creative session to update a spreadsheet or search through old notes. Novel OS also includes a story debugger, an on-demand analysis tool that surfaces continuity errors, unresolved questions, and contradictions in the world you've built, so you catch plot holes before they become problems. Under the hood, the architecture is built on Next.js 15, PostgreSQL, Drizzle ORM, and AWS SQS for async processing. Every external provider, AssemblyAI for speech, OpenAI for extraction, the database, the job queue, sits behind a clean interface, making the system easy to extend and swap without touching application logic. Novel OS turns your voice into a structured, queryable story world so you can write faster, think bigger, and never lose the thread.
30 Sep 2026