Case Studies that ship.

Case Study 01

Turning AI Into a Co-pilot Auditors Trust

Reframing a black-box GenAI tool inside a Big Four risk platform — designing for transparency, recoverability, and the auditor's mental model of evidence.

Problem Auditors couldn't trust AI-generated evidence — no transparency, no control.

Solution Redesigned the GenAI layer with evidence trails, confidence indicators, and override flows.

Outcome Adopted across EY risk teams; GenAI reframed from liability to trusted co-pilot.

GenAI · Risk Trust & Transparency

Read case study →

Case Study 02

Embedding AI Into the Workflow, Not Beside It

A contextual AI widget built to live inside EY's risk platform — bringing AI to the exact moment auditors need it, without breaking their workflow or trust.

Problem Auditors valued GenIAus but resented leaving their risk dashboard to use a separate AI tool.

Solution Designed Genie — a layered contextual widget embedded in the VIA platform with page awareness and deep workflow integration.

Outcome −60% context-switching; time-to-first-query dropped from 4 min to ~45s; +28% trust rating.

GenAI · EY Embedded AI

Read case study →

Case Study 03

Empowering Farmers Through Sustainable Practice

A research-led AgriTech tool for the World Bank that became EY's first engagement considered for patent application — translating field research into a mobile-first system.

Problem Smallholder farmers had no accessible tool to adopt sustainable practices at scale.

Solution Field-research-led mobile tool built around real farmer workflows — simple, vernacular-ready.

Outcome EY's first patent-considered engagement; validated by the World Bank as a replicable model.

World Bank · AgriTech Patent-pending

Read case study →

Case Study 04

Civic Complaint Isn't a UX Problem. It's a Trust Problem.

A research-led redesign of citizen complaint infrastructure for Indian municipalities — starting not with the form, but with the structural failure of trust that made citizens stop raising issues.

Problem Citizens stopped raising complaints because nothing got resolved — no transparency, no accountability, no feedback loop.

Solution Designed ResolveNow to attack four broken mental-model beliefs: visible officers, real-time tracking, ward-level accountability, and a closed feedback loop.

Outcome A system designed to repair civic trust — not just digitise a form — validated through multi-lens research across citizens and municipal officials.

Civic Tech · E-Governance India

Read case study →

Case Study 05

A Dashboard Shows What Happened. A Digital Twin Shows What Could.

RAP — EY's AI-assisted digital twin canvas for risk analytics. What started as a multi-tool integration became a probabilistic simulation engine steered through conversation.

Problem Analytics managers navigated 6+ fragmented risk tools — VIA, IRL, Symphony, Nexus — manually compiling data and losing context across every switch.

Solution Designed a multi-pane canvas that holds tools simultaneously, evolving into an AI-driven digital twin with conversational steering for four distinct personas.

Outcome One engine serving CRO, Internal Auditor, Risk Consultant, and Controls Manager — with uncertainty visualisation and what-if scenario simulation.

EY · Enterprise GenAI Digital Twin · Risk Analytics

Read case study →