Designing for scientists who know exactly what they want.

Most design work involves closing a gap between what a product can do and what a user understands. Bioinformatics flips that - research scientists understand the domain at a level that's genuinely humbling, and they have strong opinions about how their tools should behave. The usual instinct to simplify is often exactly wrong: oversimplify a genomic data visualization and you've broken the thing that makes it useful. The challenge wasn't accessibility. It was credibility.

Before wireframes or visual design, I spent time understanding how scientists move through their workflows - where they hit friction, and what "good" looked like in their eyes. Bioinformatics has established conventions, and violating them without good reason signals immediately that the product wasn't built for them. I worked closely with the tinybio team to pressure-test every decision; if a visualization obscured the data, they'd catch it. That feedback loop kept the design honest to the science.

The work spanned UX strategy, wireframes, high-fidelity prototypes, a scalable design system, and visual brand direction. For a platform this technically dense, each layer mattered - the design system meant consistency wasn't just aesthetic, it was functional, letting scientists build mental models that held across the platform. Bioinformatics tools have a reputation for being visually utilitarian; tinybio had an opportunity to be both scientifically rigorous and well-designed, and the visual language had to communicate both.