My Role

I worked on designing the employee-side chatbot experience, while senior UI/UX designer, Eng Jia Yi, focused on the admin (claim manager) dashboard, ensuring seamless claim review and approval.

My responsibilities included:

- Designing the chatbot interface for intuitive interactions and a frictionless claim submission process
- Aligning with engineering constraints by working with software engineer Mike Woolf to ensure feasibility
- Refining user flows based on iterative feedback from project manager Paul Selva Raj
- Conducting usability testing with subject matter experts (SMEs) over three structured sessions within a 3-week timeline

Key Design Decisions
- Conversational UI: A chatbot approach reduced cognitive load by guiding users step by step
- Smart Reminders: Auto-notifications prompted users to submit claims before deadlines
- OCR-Powered Receipt Scanning: Eliminated manual data entry and errors
- Transparent Claim Tracking: A centralized dashboard providing real-time status updates

Problem

KPMG employees struggle with an outdated, manual expense claim system that is slow, tedious, and prone to errors. Employees often working on tight schedules, find it difficult to submit claims on time, remember policy limits, and track reimbursements. This inefficiency leads to frustration and compliance risks.

The challenge: How might we use AI to make expense claims effortless, compliant, and fast for both employees and approvers within KPMG?

Solution

To solve this, we designed an AI-powered chatbot integrated into Microsoft Teams—a tool employees already use daily. The claim assistant simplifies and accelerates the reimbursement journey by:
- Sending smart reminders for pending claims
- Extracting receipt details via OCR (Object Character Recognition)
- Providing real-time policy compliance guidance
- Offering a clear, consolidated view of paid and unpaid claims
- Enabling claim managers to efficiently review, approve, and track expense claims in a dedicated dashboard

This solution ensures a seamless experience while improving compliance and efficiency.

Outcome

- Designed an AI chatbot prototype that enhances employee productivity and compliance
- Iterated through multiple rounds of usability testing to refine the UX

Since the chatbot isn’t live yet, we don’t have impact metrics, but potential success indicators include:
✅ Faster claim processing time
✅ Increased compliance with reimbursement policies
✅ Higher employee satisfaction and adoption

Employee UI

Recognition

"A huge thank you to everyone who contributed to ClaimNow! The demo today was a success—20 people from “Enterprise Singapore” shared today that they would love to have ClaimNow right now! That’s a testament to the power of what we’ve built together. In just 2-3 weeks, part-time, we turned an idea into a compelling AI-powered use case—blending AI UI/UX design, tech POC with real data, and a well-crafted story. This is exactly what it means to embrace an AI mindset—moving fast, iterating smart, and delivering impact.
 
Really proud of the team—this is the new way of working, and you’ve all shown what’s possible.

Great job, everyone!"

Talal Ikhwan, Head of KPMG Digital Village