Scaling design as a system, not a service.

Led DesignOps and system adoption across 12 product teams at SeamlessHR — operationalising design as infrastructure: how decisions are made, how knowledge compounds, and how teams stay coherent as the organisation scales. Principles built here are designed to hold beyond any single product organisation.

12

Product teams aligned

8

Designers led

88.9

Adoption score /100

91.1

Trust score /100

↓30%

Onboarding time

↓50%

Token misuse

~80%

Daily usage

2.49M

Token insertions

What I believe about design at scale.

  • 01Design is infrastructure, not a service.
  • 02Systems compound. One-offs decay.
  • 03Operationalise decisions before they become bottlenecks.
  • 04Onboarding is the fastest path to system literacy.

Seven layers that hold the system together.

01

System Operationalization

Turning the design system from a library into operating infrastructure — token architecture, governance model, contribution pipeline, and release cadence.

02

Contribution System (SOP)

A defined contribution pipeline: proposal → critique → spec → review → release. Anyone in product or engineering can contribute, with clear gates instead of bottlenecks.

03

Design Onboarding System

System-first onboarding (not tool-first). New designers learn the tokens, components, and patterns before tools — contributing back into the system within their first weeks. This approach reduced onboarding time by 30% and accelerated system literacy across every new hire.

04

Governance

Token hierarchy, naming conventions, deprecation policy, accessibility baseline, and version contracts between design and engineering. Decisions are documented, not personality-driven.

05

Team Capability

Capability maps, role-gap frameworks, structured PIPs, critique rituals, and performance frameworks. Raising the floor across the whole design function — not just the top.

06

Cross-functional Alignment

Bridging design and engineering through shared standards, design-to-dev rituals, and system contracts. Reduced rework and handoff friction across 12 teams.

07

AI Layer

Defining how AI interactions show up inside enterprise workflows as reusable patterns: trust, explainability, prompt frameworks, and shared interaction models — not one-off features.

Active questions at the intersection of systems and scale.

AI in enterprise workflows without eroding trust

How do you embed AI decision-support into high-stakes HR workflows — hiring, performance, payroll — in ways teams will actually rely on? Trust, explainability, and human override are design problems, not just engineering ones.

Governance at multi-product scale

A contribution model that works for one product team starts breaking at twelve. The question isn't how to enforce standards — it's how to design governance that teams choose to use because it makes their work easier.

Design systems as infrastructure for intelligent interfaces

As AI becomes a product layer, design systems need to encode interaction patterns for AI — not just UI components. That means reusable prompt frameworks, trust states, and fallback patterns that behave consistently across products.

Related Case Study

Seamkit — the operating system underneath.

The token architecture, governance model, and adoption programme that DesignOps is built on.

Read the case study →

AI Layer

SeamlessAI — patterns, not features.

Reusable AI interaction patterns built into enterprise workflows as a system layer — not one-off integrations.

Read the case study →