Gatherly

Summary

An industry–academia project with Samsung Electronics' home appliance division. The screens and the product details are the partner's material, so they aren't published here. What I owned and how the work was done are.

4 months

Mar – Jun 2026

Team project

13

Comparative visualisation prototypes built

Output – narrowed to 6, one chosen as the default

9,557

Lines in the code prototype

Output – 30 commits

My part – data visualisation design, code prototyping, UX writing. The research and the Figma screens were the team's work.

Overview

An industry–academia project with Samsung Electronics' home appliance division (DA). It ran from March to June 2026. It was a team project, and what the team did and what I did are not the same.

• The team – field research, personas, Figma screens, the final presentation.
• Me – data visualisation design, the code prototype, the UX writing system.

Process

Research

The team ran the research together. Only the methods are listed.

• Written interviews with 7 real users – people who use the product itself. Mar 25–30, 2026.
• In-person interviews across 3 households – households of three, four and five, visited, each member interviewed separately. Mar 27–29, 2026.
• Diary study with 3 people – a woman and a man in their twenties, a woman in her forties. Mar 25–28, 2026.
• One expert interview with store observation – interviewing a sales specialist in a retail store and watching where people look at a screen first.
• Desk research – trend mapping and a comparison of competing products.

Prototype

I built the comparison tool before choosing a screen, so the evidence wouldn't be assembled after the fact.

• 13 visualisation renderers – the same dataset drawn 13 ways and compared side by side. Six were shortlisted and one became the default. All of it is output from code I wrote.
• The criteria came first – readable in a glance, magnitude legible without thinking, several values held at once, uncertainty expressible.
• Uncertainty moved into visual properties – probability as colour temperature, confidence as the stability of an edge. The goal was to stop an uncertain value from looking certain.
• UX writing – 28 strings rewritten one by one, branched by probability band into roughly 60 variants, with one rule throughout: don't use a confident voice for an uncertain value.