A shared foundation. A faster way forward.

As Anaconda brought acquired products together, our design foundations needed to help the organization move in the same direction.

Two design systems and differing backends created friction. I led UX across three acquisitions toward a connected desktop and SaaS experience, partnering with product, engineering, and executive leadership on a phased convergence plan.

The component transition was one part of that larger effort. The opportunity was to give teams a common starting point for exploring ideas, discussing decisions, and preparing work for engineering.

I pitched moving from Kendo to shadcn/ui, with a lightweight foundation that could support AI-assisted prototyping. My role was to set the direction, connect the work to the convergence strategy, and help the team turn that direction into reusable foundations.

The team built baseline components, custom patterns, and implementation guides. We connected those foundations to our internal Claude environment so prototypes reflected the product.

Kendo provides packaged components and matching themes. shadcn/ui gives the team editable component source, along with responsibility for maintenance, accessibility, and upgrades. The decision was about the way we needed to work, rather than one library being universally better.

With shared foundations connected to our internal Claude environment, company teams could produce prototypes rooted in our actual product design language. Ideas and sketches became clickable concepts sooner.

Those concepts helped stakeholders have more useful conversations before initial development handoff. We could walk through experiences, identify gaps, and refine the PRD together. AI helped make the work tangible; design judgment, research, and critique still shaped the direction.

Across the workflow from idea to concept to development-ready designs, we saw a 30% efficiency improvement, supported by quantitative measurement and qualitative feedback. Engineering received something concrete to work from roughly a week and a half to two weeks sooner.

Those improvements came from shared foundations and AI-assisted workflows together. They describe the design workflow, rather than coding time or production-release speed.

By September 2026, the baseline design system, custom components, patterns, implementation guides, and connected prototyping workflow were in place. The full integration of all three acquisitions was still underway when I left. The foundations and delivery plan were in place.

For me, leading this work meant giving the team a clear direction and the room to execute it. The components mattered because of what they enabled: a shared starting point, earlier conversations, and more concrete decisions.