Making customer support work way easier with generative AI

Making customer support work way easier with generative AI

Making customer support work way easier with generative AI

Project Overview

Project

scope

Enterprise UX has a bright future ahead introducing Artificial Intelligence into its service flows.


Rogers has adopt it in a pilot project, supporting agents with real time data, context and answers — supporting customers faster and more precisely.

Project

scope

Enterprise UX has a bright future ahead introducing Artificial Intelligence into its service flows.


Rogers has adopt it in a pilot project, supporting agents with real time data, context and answers — supporting customers faster and more precisely.

Design process

1

Partnered with PO to help design vision based on initial pilot constraints with Open AI API.

2

Held sessions with business stakeholders to align vision and user experience.

3

Socialized solutions to design and leadership teams.

4

Led the effort to guarantee quality, performing constant UX/UI testing.

Design process

1

Partnered with PO to help design vision based on initial pilot constraints with Open AI API.

2

Held sessions with business stakeholders to align vision and user experience.

3

Socialized solutions to design and leadership teams.

4

Led the effort to guarantee quality, performing constant UX/UI testing.

Design team management

1

Led a team of 2 designers

2

Managed design roadmap, creating and prioritizing design stories.

3

Facilitated daily sessions for review and collaboration.

Design team management

1

Led a team of 2 designers

2

Managed design roadmap, creating and prioritizing design stories.

3

Facilitated daily sessions for review and collaboration.

Business impact

1

Agent Assist was deployed initially to Live Chat tool, helping xxxxx agents.

2

First ever AI project for Rogers, responsible for gathering more budget for future enterprise UX improvements.

3

On average, agents are reducing chat response time in XX%.

Business impact

1

Agent Assist was deployed initially to Live Chat tool, helping xxxxx agents.

2

First ever AI project for Rogers, responsible for gathering more budget for future enterprise UX improvements.

3

On average, agents are reducing chat response time in XX%.

Key challenges

1

3rd party tool limitation on customizing the UX/UI.

2

Navigating AI uncertainty which led to constant requirement changes.

3

Timelines were tight. Inception to pilot launch in 4 months.

Key challenges

1

3rd party tool limitation on customizing the UX/UI.

2

Navigating AI uncertainty which led to constant requirement changes.

3

Timelines were tight. Inception to pilot launch in 4 months.

Design

impact

1

First project at Rogers to define UX/UI guidelines for AI initiatives.

2

Introduced

3

More than 20,000 agents use OneView.

4

Wireless transactions accounted for $10 billion dollars in 2023.

Design

impact

1

First project at Rogers to define UX/UI guidelines for AI initiatives.

2

Introduced

3

More than 20,000 agents use OneView.

4

Wireless transactions accounted for $10 billion dollars in 2023.

Project timeline

1

Jan – Proof of concept and initial design vision.

2

Feb to Mar – MVP Design ideation + development.

3

Apr – MVP Launch.

Project timeline

1

Jan – Proof of concept and initial design vision.

2

Feb to Mar – MVP Design ideation + development.

3

Apr – MVP Launch.

Design in numbers

1

All 4 major flows redesigned: new customers, upgrading family or single plan and phones

2

On average, each flow has 50+ pages, excluding responsive designs.

3

This project includes thousands of insertions from Rogers design system library in Figma.

Design in numbers

1

All 4 major flows redesigned: new customers, upgrading family or single plan and phones

2

On average, each flow has 50+ pages, excluding responsive designs.

3

This project includes thousands of insertions from Rogers design system library in Figma.

Next

steps

1

Support frontline team on MVP launch in October/November.

2

Improve solutions based on initial feedback gathered from MVP launch.

3

Document UI and interactions for other AI projects to use design component library.

Next

steps

1

Support frontline team on MVP launch in October/November.

2

Improve solutions based on initial feedback gathered from MVP launch.

3

Document UI and interactions for other AI projects to use design component library.

Problem statement

Initial assumptions

Goal: redesign the retail system based on the system used in call centres.

Our initial assumption is that the current retail system was outdated, slow and overly complex – hurting sales, agent experience and effectiveness.

Rogers has another customer channel through the call centres, which utilizes a complete different system.

Based on internal metrics regarding transaction times and success rate, the call centre system has over performed the retail system on many aspects, specially user feedback.

Pain #1

Being a legacy system, it lacks speed and connectivity with other complex systems like knowledge base.

Pain #2

Pricing and offers are better exposed in other channels, making difficult to close sales.

Pain #3

Overall sale flows are out of order, making agents start flows from scratch multiple times.

Why addressing

the problem

Business perspective: A potential increase in revenue due to the alignment of multiple sale channels through one sale/support system.

Design perspective: Making the experience more universal, designing for consistency and scalability.

Agent perspective: Helping agent journeys the easiest way possible, where they don't have to keep relearning patterns and functions.

Design process

1

Partnered with UX researches and store agents to analyze and identify pain points in the current system.

2

Held sessions with business stakeholders to align vision and user experience.

3

Socialized solutions to design and leadership teams.

4

Led the effort to guarantee quality, performing constant UX/UI testing.

Design process

1

Partnered with UX researches and store agents to analyze and identify pain points in the current system.

2

Held sessions with business stakeholders to align vision and user experience.

3

Socialized solutions to design and leadership teams.

4

Led the effort to guarantee quality, performing constant UX/UI testing.

Design process

1

Partnered with UX researches and store agents to analyze and identify pain points in the current system.

2

Held sessions with business stakeholders to align vision and user experience.

3

Socialized solutions to design and leadership teams.

4

Led the effort to guarantee quality, performing constant UX/UI testing.

Design process

1

Partnered with UX researches and store agents to analyze and identify pain points in the current system.

2

Held sessions with business stakeholders to align vision and user experience.

3

Socialized solutions to design and leadership teams.

4

Led the effort to guarantee quality, performing constant UX/UI testing.

Design process

1

Partnered with UX researches and store agents to analyze and identify pain points in the current system.

2

Held sessions with business stakeholders to align vision and user experience.

3

Socialized solutions to design and leadership teams.

4

Led the effort to guarantee quality, performing constant UX/UI testing.

Design process

1

Partnered with UX researches and store agents to analyze and identify pain points in the current system.

2

Held sessions with business stakeholders to align vision and user experience.

3

Socialized solutions to design and leadership teams.

4

Led the effort to guarantee quality, performing constant UX/UI testing.

Research and discovery

Initial assumptions

Goal: redesign the retail system based on the system used in call centres.

Our initial assumption is that the current retail system was outdated, slow and overly complex – hurting sales, agent experience and effectiveness.

Rogers has another customer channel through the call centres, which utilizes a complete different system.

Based on internal metrics regarding transaction times and success rate, the call centre system has over performed the retail system on many aspects, specially user feedback.

Goal: redesign the retail system based on the system used in call centres.

Our initial assumption is that the current retail system was outdated, slow and overly complex – hurting sales, agent experience and effectiveness.

Rogers has another customer channel through the call centres, which utilizes a complete different system.

Based on internal metrics regarding transaction times and success rate, the call centre system has over performed the retail system on many aspects, specially user feedback.

Goal: redesign the retail system based on the system used in call centres.

Our initial assumption is that the current retail system was outdated, slow and overly complex – hurting sales, agent experience and effectiveness.

Rogers has another customer channel through the call centres, which utilizes a complete different system.

Based on internal metrics regarding transaction times and success rate, the call centre system has over performed the retail system on many aspects, specially user feedback.

Pain #1

Being a legacy system, it lacks speed and connectivity with other complex systems like knowledge base.

Pain #2

Pricing and offers are better exposed in other channels, making difficult to close sales.

Pain #3

Overall sale flows are out of order, making agents start flows from scratch multiple times.

Why addressing the problem

Business perspective: A potential increase in revenue due to the alignment of multiple sale channels through one sale/support system.

Design perspective: Making the experience more universal, designing for consistency and scalability.

Agent perspective: Helping agent journeys the easiest way possible, where they don't have to keep relearning patterns and functions.