Skip to content

Professional

NDA

Evaluating Design-to-Code Tools

UX Researcher & Business Analyst

Mar – Sep 2025, Bachelor Thesis at BMW

For my BMW thesis, I screened 29 tools and developed a business case estimating potential implementation time savings of up to 61% with design-to-code tools.

I wanted to find out whether design-to-code could work in BMW’s frontend environment. I compared the market and looked closely at developer workflows before building an enterprise proof of concept.

This project is under NDA, so I describe it in general terms.

The short version

  1. Started with 29 tools, narrowed them to 11 and shortlisted two before testing the strongest candidate
  2. Surveyed 20 frontend developers about handoff tasks, time spent and automation needs
  3. Presented the AI4UX evaluation to the BMW Group UX Live Center and earned a 1.0 thesis grade

What I worked on

  • I compared selected tools using a shared component setup and weighted criteria for code quality, Angular support, design fidelity and integration effort.
  • I surveyed developers about concrete handoff tasks and time spent, then used their answers with the benchmark to build the business case.

What came out of it

The business case estimated potential implementation time savings of up to 61% using design-to-code tools compared with manual implementation. Maintainability, design-system conformance and rework remained critical to the adoption decision.

What I learned

The benchmark became much more useful once I matched it to the way teams actually work. Asking developers about specific tasks gave me clearer answers. Discussing effort and risk made the evidence easier for stakeholders to act on.

Tools and methods

  • User Research
  • Survey Design
  • Benchmarking
  • Business Case Development
  • Figma
Back to overview