Case Study

Design System

The Cloudaware Design System is the largest and longest-running initiative I have led at the company. It touched not one product or one team, but the entire visual and technical foundation on which both marketing and the product itself were built.

The problem

Cloudaware CMDB had been assembled by hand by different developers for years, with no shared visual language.

There were not enough components, and some screens were built almost from scratch for a specific task
The link between design and engineering rested on one-off agreements rather than on a real process
Marketing was in the same state — the website and internal materials existed as separate entities, assembled from different elements that were never reused anywhere

How I solved it

Audit & foundations

I initiated the project once I saw the scale of the problem and started with a full audit of existing components, both in the product and in marketing. On that basis I built the core of the Design System: tokens, colors, grids and spacing.

Design–engineering process

I improved the process between design and engineering across the board: one-off agreements gave way to a single, clear process shared by product and marketing.

Starting with marketing

The rollout began with marketing: a website redesign was under way, which made it the most logical entry point. On top of the DS core I built the first set of components, and I taught the marketing designer how they are structured and how to work with them, so that interfaces could later be assembled independently and structurally correctly on top of the system.

From the website to product onboarding

The website redesign built on these components showed clearly that the system works and gave us the foundation for the next step — onboarding in the product. There, the components were not simply reused: the library was scaled and developed further, and the redesign was built on top of it. This sped up both the onboarding redesign itself and the development around it.

Adoption & AI-accelerated workflow

The hardest stage was rolling the DS out into the main product: it had been changed by different people over the years under different technical constraints, so the work had to be done carefully and in stages — fonts first, then colors, spacing and so on. I did most of this work together with AI: I worked closely with VS Code and Claude Code, which sped up the rollout and reduced the risk of errors.

Today the core layer is about 70% implemented: we are cleaning up tokens and gradually scaling and refining components, starting with the key layout components. It is still an active, long process, but the infrastructure and processes are already in place and working.

Separately, within this project I mentored a designer from the marketing team for a year: I taught her the fundamentals of the Design System and the principles of working with it, how to build and correctly structure components and layouts, how to organize and maintain Figma files, and how to integrate AI into design, and helped her grow professionally.