Election certification at platform scale

Rather than centralizing all design work in a single backlog, I embedded in each team's rhythm — attending standups, reviewing sprint commitments, and translating certification requirements into actionable UX constraints without becoming a bottleneck on any one surface.

The trade-off was visibility: I would never be the primary designer on every screen, but I could ensure that what shipped across teams read as one platform. That model only worked if I documented decisions in a format engineering could reuse without me in the room.

Certification documents are usually treated as compliance checklists downstream from design. I reframed them as input to information architecture — if a jurisdiction auditor needs to reconstruct a decision trail, the UI has to expose that trail without burying operators in audit chrome during routine tasks.

That shift changed how product and engineering prioritized backlog items. Features that looked cosmetic often carried certification risk; features that looked critical sometimes didn't. Design became the translation layer between regulatory language and sprint-ready acceptance criteria.

Early in the engagement, a full design-system migration looked like the fastest path to consistency. Discovery with engineering revealed that a wholesale swap would stall certification on multiple teams simultaneously — the exact outcome the business could least afford.

I redirected the effort toward incremental pattern adoption: shared tokens and documentation standards first, component migration sequenced by team readiness. The migration didn't ship in its original form, but the redirect kept certification timelines intact while still moving the platform toward coherence.

With NDA constraints limiting what could be shared externally, internal documentation became the primary artifact proving design value. I established a working format — decision logs, pattern rationale, before/after workflow maps — that product managers could attach to roadmap proposals without a design review meeting.

The format spread because it reduced handoff cycles: engineering could implement from documented intent, and product leadership could evaluate proposals against North Star principles before committing sprint capacity.

Without a shared vision document, each team optimized locally. I partnered with product leadership to draft a concise North Star — not a pixel spec, but a set of principles that could reject scope before it entered a sprint.

Principles only matter if they're used. The operational test was whether PMs referenced them unprompted in planning conversations. By the second quarter, they were — which meant design influence scaled beyond my direct embedding hours.