KPMG Global Workbench: Designing the Foundation for Scalable, Trusted AI

Overview

Led design strategy for the Global AI Experience team, developing a comprehensive AI Design System with a unified UI/UX vision to enable KPMG member firms worldwide to efficiently design, build, and scale intuitive, trustworthy AI applications.

The challenge wasn't just visual consistency. It was establishing shared patterns for how AI should behave, communicate uncertainty, and hand off to humans across 250,000+ employees in Advisory, Audit, Tax, and Business Services.

The Problem

Member firms struggled to design and scale AI applications that felt intuitive and trustworthy within employee workflows. Each team had built its own patterns for AI interactions, loading states, error handling, and trust signals. The result was a fragmented experience that eroded user confidence in the AI itself.

Inconsistent UI/UX approaches across teams and regions impeded efficient scaling. The absence of standardized AI interaction guidelines increased design time and user confusion, and balancing innovation with regulatory compliance slowed timelines further.

Research

Conducted 30 in-depth interviews with KPMG employees across Advisory, Audit, Tax, and Business Services spanning Americas, EMEA, and ASPAC regions at associate-to-partner seniority levels.

Facilitated co-creation workshops including "Day in the Life" and "Create Your Own Dashboard" activities to identify workflows, surface AI opportunities, and test interaction concepts. Three consistent findings emerged: frustration with inconsistent AI experiences, hesitancy caused by opaque recommendation logic, and a strong preference for AI embedded directly into existing workflows rather than separate tools.

Design Solution

Developed a universal AI Design System with standardized UI components and interaction patterns ensuring consistency across member firms and regions. Introduced transparent AI interaction guidelines with explainability elements to clarify how recommendations are generated and build system confidence.

Created context-aware, workflow-integrated UI components enabling seamless AI tool embedding within existing applications, reducing friction and promoting adoption. The system was documented with usage guidelines that treated AI behavior, not just visual style, as a first-class concern.

Testing and Iteration

Conducted usability testing with cross-functional teams from Advisory, Audit, and Tax, validating component usability and AI interaction patterns. Facilitated iterative design reviews with global member firms, refining components based on regional feedback.

Piloted the Design System with select member firms to test regional adaptability while maintaining global consistency, collecting localization feedback across languages and regulatory contexts.

Outcome

40% reduction in design and development time for AI applications through standardized components and guidelines. Increased AI tool adoption rates across member firms, with positive feedback emphasizing improved usability and consistency.

Enhanced user trust and engagement through transparent AI interactions, resulting in measurable satisfaction score increases during pilot rollouts. Global Workbench became the shared foundation for AI product development across KPMG's enterprise portfolio.

Key Learnings

Simplify and standardize. Consistency and simplicity prove essential when designing AI experiences for diverse, global teams.

Design with users, not just for them. Regular testing and co-creation ensure regional relevance while maintaining global scalability.

Build trust through transparency. Clear explanations and user control foster confidence in AI-powered tools.