AI Assistant - Watson

Watson is our AI Assistant chat bot that can help with story plotting & research, creative inspiration, contextual analysis & feedback for your project, story, characters, collections and more. A premium subscription is required to use the AI assistant.

To use Watson, simply click the robot icon in either your project dashboard, draft editor, collection items index, or within an item itself.

Note: Watson is not sent any data until you explicitly do so through the AI assistant chat window.

Watson Analysis Contexts

The Watson assistant has several contexts in which it can be used: project, draft, collection, and collections and items. Contexts are related to what content is sent to Watson for analysis. For example, the chapter context will send only the chapter title and content to Watson for analysis. The draft context will give all chapters and scenes and their content for analysis.

Draft Context

Watson will be given all parts, chapters and scenes in your draft as context for Watson to give feedback and analysis on. You can use Watson to analyze your story pacing, character development, plotting and more.

Chapter Context

Watson will be given only the relevant chapter for analysis and feedback. This can be used if you wish to conserve token usage.

Project Context

When using the project context, it will only send Watson the "meta" information for your project, stories, drafts, collections and items. "Meta" information are things like project name and description, story titles and descriptions, chapter titles and descriptions.

Collection Context

Watson will be given the collection's label, description, the "meta" information for all the collection's items.

Collection Item Context

Watson will be given the item's name, description, content, and properties for analysis and feedback.

Tips on making the best use of Watson

• Filling out your project's name with a descriptive name and description can help Watson give more relevant feedback
• Using chapter context rather than the draft-wide context can help with conserving token usage if your draft gets very long

If you feel any of this information is not clear or missing something, please contact us for further questions.