Case study · Mantle Bio
Mantle Bio, a YC seed-stage startup, was building data infrastructure for computational biologists. I joined just after its public beta, which ran on a UI that had been built before any design work. The product was mid-pivot toward a platform model, with use cases and scope still in motion. I worked directly with the CEO and CTO, who had engineering backgrounds and who had never worked with a designer.
On paper, the job was defining the interactions. Underneath, it was helping technical founders find focus, and grounding their decisions in how scientists actually work. That required a shared language between design, the founders, and users.
From crossed wires to connecting with users
Created a shared language
Early meetings had a lot of miscommunication. I leaned into the technical side of design: hierarchy, architecture, flows.
We connected on their terms, and product work accelerated.
Focused the room
I built collaboration around decision-making: is something standalone, nested, or part of something that exists?
Nothing was added by default, and a clearer product began to take shape.
Committed to the core
Once exploration was over, I explicitly designed the core workflows to be foundational and said so.
The founders kept the core intact through every later change.
Handed them the users
I designed the user study with instructions for how to read responses past “it’s fine.”
The founders led the sessions and got to experience users’ thinking firsthand.
With new collaborators, the first thing I design is how we make decisions together. The founders largely deferred to me on UX, so I used that trust to create focus. In meetings, three questions turned open discussion into a decision about where each new idea should go, and framing the navigation as the product backbone meant that new ideas had to earn a place rather than default into the global nav.
The language gap I’d first hit with the founders showed up with users, too. Testing prototypes showed that the product’s engineering terms didn’t land with scientists, and neither did my icon-only controls. I updated the interface to have plain labels and clearer terms to create the start of a better vocabulary.
The throughline
Mantle’s founders needed focus more than they needed screens. Learning their language and their users’ came first, then the decision structure, then direct exposure to users. This is where I start with any new team now: learn how they think, design how they decide, then put them in front of the people they’re building for.
© 2026 Acorn Steed. All rights reserved.