/services/design

Interfaces for products that think

AI features fail in the interface more often than in the model. We design chat, suggestions and automation so users always know what the system did, why, and how to overrule it.

// what we do

Where we earn our keep

  • AI interaction design. Streaming answers, citations, confidence signals and undo. The patterns that make people trust an intelligent feature instead of fearing it.
  • UX research. We watch real users attempt real tasks and design around what they do, not what a stakeholder assumes they do.
  • UI design and design systems. Clean, fast interfaces with reusable components and rules, so your product still looks like one product two years of features later.
  • Accessibility. WCAG 2.2 AA as a floor. Keyboard navigation, contrast and screen reader support built in, including for streaming AI output.

// tools we reach for

The stack

Chosen per project, never by habit. These are the tools we most often ship with for design.

Figma Design tokens Storybook WCAG 2.2 Maze Hotjar

// deliverables

What you actually get

Every engagement ends with things you can point at, not hours on an invoice.

01

Clickable prototype

The product you can try before you pay to build it, including the AI moments, tested with users first.

02

AI interaction patterns

How the product shows sources, uncertainty, errors and the hand-off to a human. Specified, not improvised.

03

Full UI specification

Every screen, state and interaction defined, down to empty states and error messages.

04

Design system

Components, tokens and usage rules in Figma and in code.

05

Usability test findings

What real users struggled with, on video, and what we changed because of it.

06

Accessibility report

WCAG conformance checked and documented.

// next step

What could your software do if it could think?

Tell us about the work your team still does by hand. We'll reply within one working day with an honest view of what AI can and can't do for it.