Livetale is a digital TTRPG platform that uses facial recognition and voice modulation to let players fully embody their characters — voice, face, and expression. Every session is a unique performance, shaped by the players at the table.
As Senior Visual Designer, I built Spellbook — the team's design system spanning both the web platform and Unity game client. I partnered with web and Unity developers to define component specs, establish contribution workflows, and ensure design intent translated cleanly to implementation.
Identifying the problem
Roleplaying games are played partially in the imagination and partially using a game board or VTT, or in the case of larping, a space for interaction.
The connection we have to our character is limited to the voice acting, appearance, and expressions of our characters.
When telling complex stories with moments of joy, sadness, anger, and surprise, we need tools that help others feel the things we are trying to express.
Livetale is a digital platform that leverages MediaPipe for facial recognition and integrated voice modulation.
This way, even if you have a deep voice, you can create a fairy princess elf character, with long flowing hair, and utilize voice modulation to sound like a high pitched femme fatale.
Players can come together to be who they want to be - voice, face, and body - all controllable so you can play our dramatic melodramas, serious true crime, or vintage detective fantasies with new and old friends.
Research the Market
We mapped the VTT landscape — Roll20, Owlbear Rodeo, Foundry, TaleSpire, and others — looking for where Livetale could stand apart. Most competitors lean heavily on 2D top-down boards and a rugged fantasy aesthetic that skews older. We saw an opening.
Most VTTs default to the grimdark Dungeons & Dragons visual language — heavy, aged, narrow. Our audience was 18–28, and we looked to Valorant, Breath of the Wild, and Genshin Impact as north stars: cyber-fantasy with broad appeal and a fresh palette.
Designing for Users
User Flow from Main Menu to Gameplay
I utilized FigJam to create user flows for features such as User Permissions, Tale Creation, Menu Explorations, and the pipelines of Web <-> Software interaction. This put engineers and artists on the same page for how the game loop would be structured.
I built Spellbook as a fully dynamic component library — variables exposed, states animated, and specs documented for both web and Unity implementation. The system was designed to scale: developers could swap tokens and update components without touching core layouts, and every animation in Figma had a direct implementation reference.
This project sharpened my systems thinking. Building Spellbook from scratch — solo, at an early-stage startup — meant making decisions that had to hold up as the product scaled. Designing for both web and Unity simultaneously surfaced real-world constraints that made every component more considered. My goal going forward: push the visual language further into stylized game territory, with more color and character.