AI Playable Novel

Timeline

3 months

Role

Founding Designer

Skills

Product Design, UI/UX Design

Team

Entrepreneur Project

Introduction

This project explores how generative AI can enable a new form of playable narrative experience, where stories evolve through player decisions instead of fixed scripts.

The platform combines AI-generated narrative, characters, and images to create an interactive storytelling system. Players enter a story world, choose roles, and influence the narrative through their choices.

When I joined the project, the early prototype focused on character creation as the entry point. My role as a Product and UX Designer was to restructure the interaction model so that story generation, player decisions, and AI capabilities could work together as a coherent gameplay system.

My design helped the project secure $125k in funding.

Optimizing Landing Page

Strategic Challenge (Problem):

Visual Stagnation: Standard grid layouts failed to differentiate story genres or highlight high-fidelity AI artwork.
Engagement Drop-off: Lack of a clear visual anchor led to passive scrolling and lower click-through rates.

Design Execution & Impact (Solution)

Zigzag Style Layout: Implemented a rhythmic, asymmetrical structure to maximize visual impact and drive story exploration.
Dynamic Discovery: Created a curated browsing experience that improved scannability and "browse-to-play" conversion.

Reframing Onboarding: Aligning Characters with Narrative Context

Strategic Challenge (Problem):

Cognitive Friction: Users struggled to conceptualize or design original characters in a vacuum without an established story world or genre.
Narrative Disconnect: "Character-First" flows often led to personas that felt out of place within the plot, diminishing user immersion and logic.
Development Constraints: From an MVP angle, the AI model lacked the stability to integrate random characters into plots without pre-existing world logic.

Design Execution & Impact (Solution)

Story-First Architecture: Re-engineered the flow to establish world-building first, providing users with a concrete narrative anchor to lower the creative barrier.
User Mental Models: Leveraged a natural cognitive path (Environment → Identity) to ensure users understood the game's stakes before defining their role.
Structural Stability: Optimized backend performance by setting a "world-state" before character generation, significantly reducing model hallucinations.

Designing Complex Game Mechanics: Multiplayer Mode

System Architecture:

From-Scratch Framework: Designed a proprietary multiplayer cycle consisting of 12 high-energy storytelling rounds.
Dual-Information Logic: Architected a "Public Script" for collective progression alongside "Private Clues" for individual strategy.
Timed Engagement: Optimized each round to a 1.5-minute duration based on user research to maintain peak narrative momentum.

From-Scratch Framework: Designed a proprietary multiplayer cycle consisting of 12 high-energy storytelling rounds.
Dual-Information Logic: Architected a "Public Script" for collective progression alongside "Private Clues" for individual strategy.
Timed Engagement: Optimized each round to a 1.5-minute duration based on user research to maintain peak narrative momentum.

From-Scratch Framework: Designed a proprietary multiplayer cycle consisting of 12 high-energy storytelling rounds.
Dual-Information Logic: Architected a "Public Script" for collective progression alongside "Private Clues" for individual strategy.
Timed Engagement: Optimized each round to a 1.5-minute duration based on user research to maintain peak narrative momentum.

Interaction Design:

Zigzag UI Interface: Structured a split-screen layout with AI visuals/dialogue on the left and interactive options on the right.
Simultaneous Decision Model: Implemented a concurrent selection mechanic to eliminate downtime and reduce player friction.
Cross-Functional Sync: collaborated with developers to ensure real-time data synchronization between AI and multi-user inputs.