

2025
Makiverse
Makiverse
Makiverse focuses on multi-step AI image generation for consistent anime-style characters, manga panels, dialogue, and props.
AI
Entertainment
Project Overview
Designing the future of AI Story-telling.
Summary
At the intersection of AI and narrative design, I led product design for Makiverse and CRP; two interlinked platforms redefining how people create, personalize, and evolve characters and stories through generative AI. As the sole designer across both products, I was responsible for building an intuitive, scalable design system that empowered users to co-create alongside machine learning models.
The Problem
AI image generation tools were powerful, but lacked structure and consistency. Users couldn’t easily create persistent characters, manage inventory, build scenes, or control style evolution. Existing workflows were linear, opaque, and optimized for one-off outputs, not storytelling or world-building.
Key Frictions
No continuity across generations ("every output looked like a different person")
Complex systems (LoRA, tiered outputs) overwhelmed non-technical users
Lack of contextual UI made advanced features feel inaccessible
Generation > interaction loop was broken
Users couldn’t combine characters, props, and dialogue into narrative scenes
Character Generation

Design Goals
Design Goals
Design an end-to-end generation pipeline users can trust and re-use
Enable consistent characters and inventory-like props for reusability
Build tools to generate full manga panels with character placement, layout control, and dialogue
Clarify ML model behavior with transparent UX (LoRA, tier tiers, failures)
Support both novice and power users across web + mobile
Establish monetization and reward systems that align with creative goals
Design
Design
Multi-Step Generation Flow
A modular UX for generating AI images, broken into:
Prompt input
Reference image upload
LoRA training (user-selected)
Tiered generation (speed, quality, cost)
Each step was progressively disclosed, with feedback on model behavior, generation time, and expected fidelity. This helped reduce cognitive overload while increasing success rates.
Consistent Characters & Identity Management
Designed a character sheet system enabling users to create, evolve, and re-use unique characters across generations.
Persistent identity tagging (style, genre, traits)
Character portrait system embedded into the generation UI
Real-time preview of character-consistent outputs
Prop & Inventory UX
Inspired by RPG loadouts, I created a system for users to design, assign, and organize character props. Each prop was AI-generated and visually attached to a specific character.
Users click the “+ Prop” button per character to trigger a secondary generation flow
Props displayed in a horizontal scrollable toolbar tied to the selected character
Visual inventory states (active, assigned, unassigned)
Manga Panel & Dialogue Generation
Extended the image generation system to support dynamic, panel-based storytelling. Users can select from previously generated characters, assign dialogue, and prompt the model to generate full manga-style scenes.
Users select one or more characters from their roster
Input a scene prompt describing action, emotion, or environment
Add optional dialogue to guide tone or speech bubbles
System generates a full panel using AI, factoring character consistency, pose, and prop data
Checkout & Monetization Redesign
CRP featured purchasable generation tiers, diamonds, and reward mechanics. I overhauled the entire purchase flow to reduce friction.
Redesigned G2A integration for purchasing credits
Clarified diamond pack contents and tier benefits
Streamlined modal redemption flow post-purchase
Gamified Reward Systems
To increase retention and engagement, I built a badge and leveling system that rewarded creativity, consistency, and frequency.
Progression Theme:
Artistic mastery. Levels reflect creative identity (e.g. "Sketchling" to "Digital Muse")
Visual System:
Badges evolved in color, texture, and aura—culminating in a signature pink chroma at Level 100.
Consistent Item Sheet Generation


Impact
Impact
Impact
Reframed image generation from a one-off novelty into a reusable storytelling tool
Brought clarity to LoRA and inference tiers through UI metaphors and affordances
Created the foundation for a cross-platform, character-centric generative AI ecosystem
Enabled users to build full manga scenes with characters, props, layouts, and dialogue
Makiverse and CRP taught me how to shape design systems that don’t just serve function, they scaffold imagination. Designing for AI means designing for uncertainty, iteration, and trust. By architecting structured, expressive interfaces, I helped users move from generation... to creation... to narrative.
This is the future of storytelling. and I’m proud to be building it.
Manga Setup & Panel Generation


LoRA Training

More Works
(truo® — 02)
©2025
More Works
(truo® — 02)
©2025


2025
Makiverse
Makiverse focuses on multi-step AI image generation for consistent anime-style characters, manga panels, dialogue, and props.
AI
Entertainment
Project Overview
Designing the future of AI Story-telling.
Summary
At the intersection of AI and narrative design, I led product design for Makiverse and CRP; two interlinked platforms redefining how people create, personalize, and evolve characters and stories through generative AI. As the sole designer across both products, I was responsible for building an intuitive, scalable design system that empowered users to co-create alongside machine learning models.
The Problem
AI image generation tools were powerful, but lacked structure and consistency. Users couldn’t easily create persistent characters, manage inventory, build scenes, or control style evolution. Existing workflows were linear, opaque, and optimized for one-off outputs, not storytelling or world-building.
Key Frictions
No continuity across generations ("every output looked like a different person")
Complex systems (LoRA, tiered outputs) overwhelmed non-technical users
Lack of contextual UI made advanced features feel inaccessible
Generation > interaction loop was broken
Users couldn’t combine characters, props, and dialogue into narrative scenes
Character Generation

Design Goals
Design an end-to-end generation pipeline users can trust and re-use
Enable consistent characters and inventory-like props for reusability
Build tools to generate full manga panels with character placement, layout control, and dialogue
Clarify ML model behavior with transparent UX (LoRA, tier tiers, failures)
Support both novice and power users across web + mobile
Establish monetization and reward systems that align with creative goals
Design
Multi-Step Generation Flow
A modular UX for generating AI images, broken into:
Prompt input
Reference image upload
LoRA training (user-selected)
Tiered generation (speed, quality, cost)
Each step was progressively disclosed, with feedback on model behavior, generation time, and expected fidelity. This helped reduce cognitive overload while increasing success rates.
Consistent Characters & Identity Management
Designed a character sheet system enabling users to create, evolve, and re-use unique characters across generations.
Persistent identity tagging (style, genre, traits)
Character portrait system embedded into the generation UI
Real-time preview of character-consistent outputs
Prop & Inventory UX
Inspired by RPG loadouts, I created a system for users to design, assign, and organize character props. Each prop was AI-generated and visually attached to a specific character.
Users click the “+ Prop” button per character to trigger a secondary generation flow
Props displayed in a horizontal scrollable toolbar tied to the selected character
Visual inventory states (active, assigned, unassigned)
Manga Panel & Dialogue Generation
Extended the image generation system to support dynamic, panel-based storytelling. Users can select from previously generated characters, assign dialogue, and prompt the model to generate full manga-style scenes.
Users select one or more characters from their roster
Input a scene prompt describing action, emotion, or environment
Add optional dialogue to guide tone or speech bubbles
System generates a full panel using AI, factoring character consistency, pose, and prop data
Checkout & Monetization Redesign
CRP featured purchasable generation tiers, diamonds, and reward mechanics. I overhauled the entire purchase flow to reduce friction.
Redesigned G2A integration for purchasing credits
Clarified diamond pack contents and tier benefits
Streamlined modal redemption flow post-purchase
Gamified Reward Systems
To increase retention and engagement, I built a badge and leveling system that rewarded creativity, consistency, and frequency.
Progression Theme:
Artistic mastery. Levels reflect creative identity (e.g. "Sketchling" to "Digital Muse")
Visual System:
Badges evolved in color, texture, and aura—culminating in a signature pink chroma at Level 100.
Consistent Item Sheet Generation


Impact
Impact
Reframed image generation from a one-off novelty into a reusable storytelling tool
Brought clarity to LoRA and inference tiers through UI metaphors and affordances
Created the foundation for a cross-platform, character-centric generative AI ecosystem
Enabled users to build full manga scenes with characters, props, layouts, and dialogue
Makiverse and CRP taught me how to shape design systems that don’t just serve function, they scaffold imagination. Designing for AI means designing for uncertainty, iteration, and trust. By architecting structured, expressive interfaces, I helped users move from generation... to creation... to narrative.
This is the future of storytelling. and I’m proud to be building it.
Manga Setup & Panel Generation


LoRA Training

More Works
(truo® — 02)
©2025


2025
Makiverse
Makiverse focuses on multi-step AI image generation for consistent anime-style characters, manga panels, dialogue, and props.
AI
Entertainment
Project Overview
Designing the future of AI Story-telling.
Summary
At the intersection of AI and narrative design, I led product design for Makiverse and CRP; two interlinked platforms redefining how people create, personalize, and evolve characters and stories through generative AI. As the sole designer across both products, I was responsible for building an intuitive, scalable design system that empowered users to co-create alongside machine learning models.
The Problem
AI image generation tools were powerful, but lacked structure and consistency. Users couldn’t easily create persistent characters, manage inventory, build scenes, or control style evolution. Existing workflows were linear, opaque, and optimized for one-off outputs, not storytelling or world-building.
Key Frictions
No continuity across generations ("every output looked like a different person")
Complex systems (LoRA, tiered outputs) overwhelmed non-technical users
Lack of contextual UI made advanced features feel inaccessible
Generation > interaction loop was broken
Users couldn’t combine characters, props, and dialogue into narrative scenes
Character Generation

Design Goals
Design an end-to-end generation pipeline users can trust and re-use
Enable consistent characters and inventory-like props for reusability
Build tools to generate full manga panels with character placement, layout control, and dialogue
Clarify ML model behavior with transparent UX (LoRA, tier tiers, failures)
Support both novice and power users across web + mobile
Establish monetization and reward systems that align with creative goals
Design
Multi-Step Generation Flow
A modular UX for generating AI images, broken into:
Prompt input
Reference image upload
LoRA training (user-selected)
Tiered generation (speed, quality, cost)
Each step was progressively disclosed, with feedback on model behavior, generation time, and expected fidelity. This helped reduce cognitive overload while increasing success rates.
Consistent Characters & Identity Management
Designed a character sheet system enabling users to create, evolve, and re-use unique characters across generations.
Persistent identity tagging (style, genre, traits)
Character portrait system embedded into the generation UI
Real-time preview of character-consistent outputs
Prop & Inventory UX
Inspired by RPG loadouts, I created a system for users to design, assign, and organize character props. Each prop was AI-generated and visually attached to a specific character.
Users click the “+ Prop” button per character to trigger a secondary generation flow
Props displayed in a horizontal scrollable toolbar tied to the selected character
Visual inventory states (active, assigned, unassigned)
Manga Panel & Dialogue Generation
Extended the image generation system to support dynamic, panel-based storytelling. Users can select from previously generated characters, assign dialogue, and prompt the model to generate full manga-style scenes.
Users select one or more characters from their roster
Input a scene prompt describing action, emotion, or environment
Add optional dialogue to guide tone or speech bubbles
System generates a full panel using AI, factoring character consistency, pose, and prop data
Checkout & Monetization Redesign
CRP featured purchasable generation tiers, diamonds, and reward mechanics. I overhauled the entire purchase flow to reduce friction.
Redesigned G2A integration for purchasing credits
Clarified diamond pack contents and tier benefits
Streamlined modal redemption flow post-purchase
Gamified Reward Systems
To increase retention and engagement, I built a badge and leveling system that rewarded creativity, consistency, and frequency.
Progression Theme:
Artistic mastery. Levels reflect creative identity (e.g. "Sketchling" to "Digital Muse")
Visual System:
Badges evolved in color, texture, and aura—culminating in a signature pink chroma at Level 100.
Consistent Item Sheet Generation


Impact
Impact
Reframed image generation from a one-off novelty into a reusable storytelling tool
Brought clarity to LoRA and inference tiers through UI metaphors and affordances
Created the foundation for a cross-platform, character-centric generative AI ecosystem
Enabled users to build full manga scenes with characters, props, layouts, and dialogue
Makiverse and CRP taught me how to shape design systems that don’t just serve function, they scaffold imagination. Designing for AI means designing for uncertainty, iteration, and trust. By architecting structured, expressive interfaces, I helped users move from generation... to creation... to narrative.
This is the future of storytelling. and I’m proud to be building it.
Manga Setup & Panel Generation


LoRA Training

More Works
©2025