Replaced "why is my bill different?" calls with an AI answer.
Research-led AI self-serve experience that explained bill changes, reduced support demand, and improved billing confidence.
🚨 TLDR Results
- → Reduced billing support calls through a scalable, AI-powered bill comparison experience
- → Replaced a failing legacy solution to improve billing transparency and customer trust
- → Led research, design, validation, and delivery across multiple business units
Context
In MyTELUS, SMB and consumer customers often contacted support when their bill changed month to month. The product opportunity was to turn those questions into a self-serve explanation customers could trust.
Problem & Goal
Customers experienced "bill shock" due to unclear billing changes, leading to high call centre volume and low trust. A previous legacy API approach had failed and could not scale.
The goal was to replace the legacy approach with a reliable AI-powered self-serve experience that reduced calls and helped customers understand what changed.
The legacy billing experience that drove customers to call centres
Role & Scope
As Lead Product Designer, I led the project end-to-end:
- Led the project from research and scoping through to high-fidelity design
- Owned design decisions and alignment across multiple business units
- Acted as the main design partner for Product, CX, Legal, and Operations teams
Approach
User-Centred Research
Identified key pain points and redefined KPIs around bill understanding and call containment — shifting the success metric from "did they read it" to "did they understand it."
Research synthesis and reframed KPI framework
Directions We Rejected
Before landing on the integrated "What's changed" tab, we explored a standalone bill comparison page. It surfaced useful data but disconnected from the customer's actual bill — more confusing, not less.
Rejected direction — too much data, not enough context
Final Production
The final solution embedded AI-generated bill comparisons directly into the billing flow — surfaced contextually, with plain-language explanations and dispute options in one place.
Final production UI — desktop and mobile, shipped October 2025
AI Quality Assurance
Led continuous QA of the AI language engine to resolve data gaps and improve clarity and tone — ensuring AI-generated explanations were accurate, plain-language, and suitable for customer-facing use.
Strategic Alignment
Worked closely with Product Owners to manage competing CX priorities and ran walkthroughs and presentations to secure approval from Business, Legal, and Operations stakeholders.
Outcome & Impact
Reduced billing-related inquiries by shifting customers to digital self-service. AI-generated explanations and digital dispute options helped customers resolve bill questions without calling support.
🌐 Media coverage via Mobile Syrup ↗
Learnings
Navigating ambiguity through structured discovery
When faced with an ambiguous problem space, I start by grounding the team in evidence — secondary research, past studies, support drivers, and available data.
That research base helps define stronger hypotheses, better KPIs, and a clearer path for design decisions before the team commits to a solution.
For AI self-serve work, the design challenge is not just the interface. It is defining what customers need to understand, how success is measured, and where human support still matters.