2025
Palette U
A website to record your memories and store them in your own 3D memory palace.
- Timeline
- Oct 2025 – Jan 2026
- Team
- Personal project
- Role
- UI/UX Designer, Project Manager, Full-Stack Developer & Marketing
- Skills
- Three.js, HTML/CSS, JavaScript, Blender, Figma, Express.js, SQL
Overview
The final project for NYU's Networked Media and Intro to Three.js courses, later extended into a complete personal project — my first Three.js and backend development project.
My roles and responsibilities:
- UI/UX design & 3D modeling — the visual design and UI/UX of the website and logo, and 3D models of the memory islands in Blender.
- Frontend & Three.js — vanilla JavaScript and Three.js for the 3D visualization and interaction; Express.js, HTML, and CSS for the website.
- Backend — Express.js and MySQL to store user data and the memory islands.
- Research & marketing — user and market research to understand needs and pain points, and marketing materials that brought the project 2,000+ views.
How can we store memories in a more convenient and meaningful way?
Target audience: teenagers, students, young adults — people who love journaling or keeping a diary.
Problems
In this fast-paced world, we are constantly flooded with new information and experiences — yet the moments that matter most quietly fade away.
User research
I conducted 20+ user interviews to understand how people relate to their memories and what stops them from preserving them.
Memory is relational.
- Memories are most meaningful when tied to people and relationships.
- Users want to revisit and share memories with others.
- Small triggers (a song, a drink) spark a whole wave of nostalgia.
Context and organization matter.
- Users think in categories — by person, place, or life stage.
- A timeline view helps users navigate rather than just scroll.
- Linked memories let users see the full picture of a moment.
Friction kills the habit.
- People are too busy for lengthy journaling rituals.
- The best capture methods are fast and low-effort.
- Even a single random photo can revive a full memory.
Market research
I researched existing memory-related apps.
Conclusion: the product should provide a memory-palace experience that enables effortless storage while keeping a clear and meaningful organization.
Visual design
