Designing the design pipeline

An ERP suite is five products pretending to be one. CRM, project management, supply chain — each module has its own workflows and its own team, serving multiple brands on web and mobile. Left alone, each one drifts toward its own buttons, its own spacing, its own opinions. And the most expensive step in the whole process was the seam between design and engineering: every screen handed off was a conversation, and every conversation was an opportunity to lose fidelity.

The answer wasn't a component library, it was a hierarchy. At the top, a cross-system design language: the decisions that hold the suite together, made once. Beneath it, two comprehensive design systems (web and mobile) that inherit the language and translate it to each platform's reality, with a live Storybook as the third track where the system meets engineering. A module team never starts from a blank canvas; they start from a system that already agrees with the other four modules.

And because the product is Arabic-first, the system is too: RTL isn't a pass at the end, it's a check built into the components, the tokens, and the handoff standards. Nineteen years of designing for Arabic taught me that bolted-on RTL always shows.

The piece I care most about: the handoff between design and the design system's GitHub repo is automated. Design decisions flow into the repository engineers actually consume, not into a slide, a redline, or a meeting. The system of record for design is the same as the system of record for code, which means drift gets caught where it starts.

The harder half of the work is governance, and it's the success metric I actually track: a contribution and maintenance process that's documented, followed, and reduces conflict. Sustainability over heroics. A design system that depends on one person's vigilance is a single point of failure with good typography.

At Moneyfellows I'd inherited a design system that designers tolerated rather than used. Improving its usability raised designer satisfaction by 60% — and taught me the lesson this whole pipeline is built on: a design system is a product, and its users are the team. If using it is harder than ignoring it, it will be ignored.

Three complex modules shipped 0→1 and a Mobile App v2 delivered — all visually and behaviorally coherent, by construction rather than by review. The pipeline is also why a team of seven can cover five modules: consistency is handled by the system, so people are spent on problems instead of patrol.