Case study
LIE Studio AI
A MacOS tool that makes AI prototyping and design incredibly predictable.
- Role
- Founder, designer and engineer
- Period
- 2023 - Present
- Context
- Solidarisoft Innovations
- Stack
- Electron · React · Tailwind
3×
Faster to working prototype
Problem
The problem
Open-ended AI generation makes prototyping unpredictable - the same prompt can yield a different result each time, which is a hard thing to build a design practice on. And the handoff from design to engineering stays a bottleneck for designers who can't code their own ideas.
Approach
How it got built
Self-taught front-end development in React, Next.js and Electron, specifically to remove that handoff bottleneck - the tool ships interactive prototypes directly to users for testing, no engineer in the loop.
A custom wireframe framework at the core: it converts wireframes into working prototypes through a defined, repeatable process rather than an open-ended generation call, which is what makes the output predictable.
Active design partnerships with service companies, startups, individual builders and founders, using their real work to validate the workflow and iterate the product.
Shipped
What I shipped
- LIE Studio AI - a macOS tool for AI prototyping and design, live at liestudioai.com.
- An AI-native design system, open source and built specifically for AI consumption.
Outcome
What it moved
- 3× faster time-to-working-prototype, with 100% predictable outputs versus open-ended AI generation.
TODO - verify
- The resume doesn't detail how the 3× and 100% figures were measured (baseline, sample, method) - verify before quoting them as a formal benchmark rather than a directional claim.
Next case study