EY's auditors had access to 80+ digital products and used almost none of them to complete their actual work. In 2022, EY partnered with Code and Theory to design a unified AI-assisted audit workspace that replaced fragmented, form-based tools with a continuous, judgment-first workflow.
Creating an efficient workspace
Auditors use various tools and information sources to ensure accurate evaluations of an entity’s financial health and adherence to legal and industry standards. As Senior UX Designer on the engagement, I was responsible for experience design across the core audit workspace, the AI support panel, and design iteration cycles. Working alongside product strategy and user research, the goal was to consolidate fragmented tools into a single, AI-assisted workspace where auditors could focus on judgment rather than logistics.
The outcome was a split-screen product that combined existing tools with AI-generated support, reducing the time spent searching for materials and references, thereby enabling auditors to focus more on decision-making during assessments.
Details
2022 - 2023
Code and Theory
Role
Concepting
User Experience
Interaction Design
Prototyping
The Challenge
EY's auditors faced three compounding problems. There was no unified process: 600+ audit activities existed across disconnected systems with no coherent flow. Legacy interfaces meant tools didn't meet modern product expectations, so auditors defaulted to Word documents. And data friction made accessing client data inside the tools what one Senior Manager described as "an absolute nightmare." The result was a workforce with access to sophisticated tooling that had essentially stopped using it.
UX Phase
Before designing, we conducted 10 to 15 in-depth interviews with Senior and Staff Auditors across geographies and engagement types. We learned that auditors orient around risk and judgment, not tasks; that guidance and documentation lived in separate places, forcing constant context-switching; and that workarounds like Word, email, and spreadsheets had become the de facto system.
From those findings, I led design exploration across the core workspace, focusing on three principles: organizing work by risk rather than task, embedding resources in context rather than in a separate tab, and replacing form-based interactions with a continuous, flow-oriented workspace.
Visual Design
We leveraged EY’s existing design system and brand guidelines to create our design comps and prototypes. The UX and Visual Design teams collaborated closely to develop a system that used EY’s branded yellow to represent guidance and AI recommendations. This partnership ensured the designs felt native to EY's existing system while pushing it forward to meet the demands of a complex, AI-assisted product.
Drag and drop interaction
A key design challenge was making the transition from AI suggestion to confirmed evidence feel intentional without creating friction. We gave auditors the ability to drag references from the support panel on the right into their documentation cards on the left. The motion was simple enough to be fast, but deliberate enough to reinforce the weight of marking something as audit evidence.
Split screen paradigm
Research made clear that the biggest source of friction was auditors constantly leaving their primary tool to access references and source materials. We reframed the problem simply: auditors need to provide answers and document evidence. That clarity drove the foundational structure of the interface: work on the left, support on the right. The split-screen paradigm wasn't just a layout choice; it was a direct response to how auditors actually think and work.
Support Panel
The support panel was designed with four distinct sections: AI suggestions for evidence based on client data, EY process methodology and knowledge, a personal file cabinet for saving materials, and a resource center to search content across EY products. Rather than requiring auditors to hunt across systems, the right context surfaced automatically based on where they were in the workflow.
Outcome
The support panel architecture and drag-to-evidence interaction were both adopted into EY's MVP product, shipped and used in daily operations. In user testing, auditors responded positively to having support surfaced in context with their documentation, validating the core design direction. The workflow-first, AI-assisted model became the directional framework for EY's broader audit tool modernization, and EY continued additional AI-related design work with our team following the engagement.
More Work
