Harrison
Verry

Technology & Client Service Leader

After nearly half-a-decade of building high-performing teams and driving commercial growth, I am returning from London to Auckland.

During my time in the United Kingdom I have delivered complex AI, web and application platforms for Apple, Amazon and Verizon across four continents.

131%
Account growth
25
Person team
15+
Global markets
Harrison Verry
Amazon· Apple· Verizon· BMW· Volkswagen· Subaru· Alexa· Dyson· Bloomsbury· Amazon· Apple· Verizon· BMW· Volkswagen· Subaru· Alexa· Dyson· Bloomsbury·

What I believe.

The best thing you can be on a complex project, with a demanding client, and a team under pressure, is the grown-up in the room. Present when it's hard, direct when it matters, and steady when everything else isn't.

Great work doesn't come from great pitches. It comes from teams that feel supported and clients that feel trusted. My job is to create the conditions for both: protecting the work, making the hard calls, and having the conversations nobody else wants to have.

I've never believed in selling the next thing before you've delivered the last. The clients I've built the deepest relationships with, including some of the largest companies in the world, came from a simple discipline: do what you said you'd do, do it well, and make the person who hired you look good. Everything else follows.

Finance isn't admin. It's a leadership responsibility. A well-run account is a healthy one, and clarity on the numbers is how you earn the trust to do bigger, bolder work.

And finally, know your technology. In a world full of AI buzzwords and overpromised platforms, the most important thing a technology leader can do is understand what they're actually building. Not to write the code, but to never be mistaken for someone selling something they don't understand.

04 roles
2022 - Present
ELVIS London
London, United Kingdom
Business Director, Technology
Previously Senior Account Director

Brought in to lead the Creative Origination arm of the Apple account before expanding to Channel Interactive (131% growth) and Business and Education (132% growth). Following the acquisition of Vixen Labs, also leads the Amazon and Verizon accounts across a team of up to 25.

  • Deliver AI platforms, web products, and voice applications for the world's largest consumer technology company across 15+ markets.
  • Grew key technology account monthly revenue by 131% through targeted commercial strategy.
  • Transformed team productivity from 53% to 99% in under a year.
  • Trusted advisor to Amazon C-suite, presenting at quarterly business reviews.
  • Reports directly to CEO and CFO.
2020 - 2022
dentsu / Isobar
Auckland, New Zealand
Client Engagement Director
Previously Client Engagement Lead · Customer Success Manager

Led all business development and client relationships for dentsu Aotearoa's Design and Experience team - delivering digital strategies, platforms and applications.

  • Drove 300% increase in team profitability from FY2020 to FY2021.
  • Onboarded Ministry of Social Development, Subaru, and Silver Fern Farms.
  • Delivered a family violence digital platform, a UK/AU/NZ e-learning platform, and a petrol company app.
2015 - 2020
Journey Digital
Auckland, New Zealand
Senior Account Manager
Previously Account Manager · Junior Account Manager

Created and sold digital strategies for leading New Zealand businesses including BMW, Volkswagen, and Fletcher Building.

  • Grew Giltrap Group retainer from $5,000 to $25,000/month.
  • Managed delivery of websites, software platforms and app builds across design, development and QA teams.
2013 - 2014
ANZ Bank
Auckland, New Zealand
Customer Services Manager
Previously Bank Teller

Client-facing roles across four of Auckland's largest branches, managing in-branch customer engagement and services.

08 areas
AI Platform Development
Delivery & strategy
Enterprise Technology
End-to-end solutions
Web, App & Voice
Alexa skills & beyond
P&L Management
Revenue forecasting
Team Leadership
25-person multi-disciplinary teams
Global Client Delivery
15+ markets
CX / UX Strategy
Audience-led design
New Business Development
Pipeline & pitching
04 studies
Verizon x ELVIS London
Conversational AI Platform
Generative AI Enterprise Conversational UX
Amazon x ELVIS London
Alexa Brand Experiences
Generative AI Voice AI LLM Architecture
Apple x ELVIS London
Help Me Choose
UX & Product Web Experience AI Recommendation
Dyson x ELVIS London
Dyson Care
Consumer Electronics Voice AI IoT
🔒
Password Protected
These case studies are confidential. Enter the password to view.
Incorrect password. Please try again.
Verizon x ELVIS London
Conversational AI Platform
Generative AI Enterprise Conversational UX

We worked with Verizon to develop sandbox solutions that enhance conversational flows across its digital touchpoints, leveraging advanced AI capabilities. The collaboration focuses on improving customer experiences through AI-powered conversation management, personalised user journeys, and intelligent access to technical knowledge bases - demonstrating our expertise in transforming enterprise-scale customer service systems with generative AI.

AI-Enhanced Utterance Detection & Deflection Reduction
We integrated AI to enhance existing conversational flows with improved utterance detection, reducing customer deflections by better understanding a diverse set of phrasings and contexts. The AI layer intelligently interprets queries, even when expressed imprecisely, and routes them to appropriate solutions, decreasing frustration and support escalations while improving first-contact resolution rates.
Personalised Conversational Flows
We developed systems that tailor conversational experiences based on individual user profiles and customer history. The AI dynamically adjusts conversation paths, recommendations, and support options based on each customer's service plan, device history, and usage patterns, ensuring relevant, efficient interactions that feel genuinely tailored.
LLM-Agnostic Architecture & Efficacy Testing
We built the system with an LLM-agnostic architecture, ensuring Verizon is not locked into a single AI provider and can leverage the best models as technology evolves. Our approach includes emulating conversations to systematically rate efficacy, identify improvement opportunities, and refine knowledge base responses - ensuring continuous optimisation and measurable quality improvements over time.
Amazon x ELVIS London
Alexa Brand Experiences
Consumer Electronics Generative AI IoT Device Experience LLM Architecture

We work directly with the Amazon Audio Ads team to deliver custom Brand Experiences via Alexa for major consumer brands. These voice-first experiences create engaging, informative interactions that connect users with brands and their products through natural conversation, whilst maintaining brand control and accuracy. Our work with Amazon represents the evolution of rule-based voice experiences into LLM-powered conversational systems.

LLM-Powered Conversational Architecture
We have been working to integrate large language models into Alexa Skills to enable users to ask any question about a brand or product and receive accurate, conversational responses. Rather than allowing unrestricted LLM output, which risks hallucinations and off-brand messaging, we have developed a hybrid approach that combines LLM natural language understanding with curated response systems, maintaining absolute brand safety while delivering the conversational experience users expect.
Curated Brand Knowledge Base & Response Templates
We have built comprehensive brand knowledge bases containing hundreds or thousands of pre-approved response templates and information slots. The LLM intelligently selects from this curated content library based on user queries, delivering information in a natural, conversational manner - eliminating hallucinations and ensuring every response aligns with brand guidelines, regulatory requirements, and approved product information.
Scalable Multi-Brand Deployment
Our solution architecture is designed for scale, enabling rapid deployment of LLM-enhanced experiences across multiple brands and product categories. We have refined processes for brand knowledge curation, response template development, and quality assurance that ensure consistent, high-quality conversational experiences while maintaining each brand's unique voice and personality.
Apple x ELVIS London
Help Me Choose
UX & Product Web Experience AI Recommendation Press Coverage

We built Help Me Choose, an interactive Mac recommendation tool for Apple.com, designed to help new-to-Mac customers find the perfect device through a short, conversational questionnaire. The experience needed to be instantly understandable, require minimal handholding, and feel tactile and natural - all within a 20-second attention window. The tool launched to widespread press coverage and social buzz.

Conversational Recommendation Flow
We designed a short, conversational flow - "Find your Mac match" - that asks users simple, jargon-free questions about how they work and what they do. The algorithm beneath it accounts for use case, budget flexibility, and performance requirements, recommending specific configurations rather than just base models. The UI was built to be evergreen, scalable as a design system, and deployable across multiple Apple web surfaces, including dedicated landing pages, family pages, and product description pages.
Intelligent Recommendation Engine
The recommendation engine was validated through three parallel streams: Mac Experts reviewed accuracy; User Testing IDIs assessed desirability, usability, and content models; and machine testing validated product distribution across all possible Q&A configurations, ensuring that no combination produced a wrong or misleading recommendation.
Press & Public Response
The launch generated coverage in Forbes, TechRadar, and Techloy, with reviewers praising the tool's brevity, plain language, and intelligent recommendations. Social response on X and MacRumors reflected genuine user satisfaction - and calls for Apple to expand the tool to iPhones, iPads, and Apple Watch.
Dyson x ELVIS London
Dyson Care
Consumer Electronics Voice AI IoT Device Experience Award Winning

We developed Dyson Care, an Alexa Voice Assistant that helps Dyson cordless vacuum owners troubleshoot problems and maintain their products through natural conversation. Designed to deflect customer service volume while enhancing the support experience, Dyson Care provides 24/7 assistance and positions Dyson as the first home appliance brand to offer conversational customer service. The project won Best Use of Audio at the 2023 Digital Impact Awards.

Voice-First Troubleshooting & Diagnostic Flows
We designed multi-turn diagnostic conversations that guide users through problem-solving via natural dialogue. Users describe an issue, and Dyson Care asks clarifying questions, provides step-by-step troubleshooting, and delivers solutions conversationally, understanding intent, accessing product knowledge, and translating technical information into accessible guidance.
Proactive Maintenance & Reminder Systems
Dyson Care enables users to set maintenance reminders and receive proactive performance tips, integrating Amazon's send-to-phone and Set Reminder features to bridge voice interactions with ongoing engagement. We identified this as a key differentiator, with none of Dyson's competitors offering similar levels of support at the time of launch.
Multilingual Deployment & Scalability
We launched Dyson Care across both the US and German markets, handling localisation of conversational flows, voice personas, and technical terminology. The solution architecture was designed for scalability, enabling rapid deployment while maintaining consistent quality across markets.
04 studies
Verizon x ELVIS London
Conversational AI Platform
Generative AI Enterprise Conversational UX
Amazon x ELVIS London
Alexa Brand Experiences
Generative AI Voice AI LLM Architecture
Apple x ELVIS London
Help Me Choose
UX & Product Web Experience AI Recommendation
Dyson x ELVIS London
Dyson Care
Consumer Electronics Voice AI IoT
🔒
Password Protected
These case studies are confidential. Enter the password to view.
Incorrect password. Please try again.
Verizon x ELVIS London
Conversational AI Platform
Generative AI Enterprise Conversational UX

We worked with Verizon to develop sandbox solutions that enhance conversational flows across its digital touchpoints, leveraging advanced AI capabilities. The collaboration focuses on improving customer experiences through AI-powered conversation management, personalised user journeys, and intelligent access to technical knowledge bases - demonstrating our expertise in transforming enterprise-scale customer service systems with generative AI.

AI-Enhanced Utterance Detection & Deflection Reduction
We integrated AI to enhance existing conversational flows with improved utterance detection, reducing customer deflections by better understanding a diverse set of phrasings and contexts. The AI layer intelligently interprets queries, even when expressed imprecisely, and routes them to appropriate solutions, decreasing frustration and support escalations while improving first-contact resolution rates.
Personalised Conversational Flows
We developed systems that tailor conversational experiences based on individual user profiles and customer history. The AI dynamically adjusts conversation paths, recommendations, and support options based on each customer's service plan, device history, and usage patterns, ensuring relevant, efficient interactions that feel genuinely tailored.
LLM-Agnostic Architecture & Efficacy Testing
We built the system with an LLM-agnostic architecture, ensuring Verizon is not locked into a single AI provider and can leverage the best models as technology evolves. Our approach includes emulating conversations to systematically rate efficacy, identify improvement opportunities, and refine knowledge base responses - ensuring continuous optimisation and measurable quality improvements over time.
Amazon x ELVIS London
Alexa Brand Experiences
Consumer Electronics Generative AI IoT Device Experience LLM Architecture

We work directly with the Amazon Audio Ads team to deliver custom Brand Experiences via Alexa for major consumer brands. These voice-first experiences create engaging, informative interactions that connect users with brands and their products through natural conversation, whilst maintaining brand control and accuracy. Our work with Amazon represents the evolution of rule-based voice experiences into LLM-powered conversational systems.

LLM-Powered Conversational Architecture
We have been working to integrate large language models into Alexa Skills to enable users to ask any question about a brand or product and receive accurate, conversational responses. Rather than allowing unrestricted LLM output, which risks hallucinations and off-brand messaging, we have developed a hybrid approach that combines LLM natural language understanding with curated response systems, maintaining absolute brand safety while delivering the conversational experience users expect.
Curated Brand Knowledge Base & Response Templates
We have built comprehensive brand knowledge bases containing hundreds or thousands of pre-approved response templates and information slots. The LLM intelligently selects from this curated content library based on user queries, delivering information in a natural, conversational manner - eliminating hallucinations and ensuring every response aligns with brand guidelines, regulatory requirements, and approved product information.
Scalable Multi-Brand Deployment
Our solution architecture is designed for scale, enabling rapid deployment of LLM-enhanced experiences across multiple brands and product categories. We have refined processes for brand knowledge curation, response template development, and quality assurance that ensure consistent, high-quality conversational experiences while maintaining each brand's unique voice and personality.
Apple x ELVIS London
Help Me Choose
UX & Product Web Experience AI Recommendation Press Coverage

We built Help Me Choose, an interactive Mac recommendation tool for Apple.com, designed to help new-to-Mac customers find the perfect device through a short, conversational questionnaire. The experience needed to be instantly understandable, require minimal handholding, and feel tactile and natural - all within a 20-second attention window. The tool launched to widespread press coverage and social buzz.

Conversational Recommendation Flow
We designed a short, conversational flow - "Find your Mac match" - that asks users simple, jargon-free questions about how they work and what they do. The algorithm beneath it accounts for use case, budget flexibility, and performance requirements, recommending specific configurations rather than just base models. The UI was built to be evergreen, scalable as a design system, and deployable across multiple Apple web surfaces, including dedicated landing pages, family pages, and product description pages.
Intelligent Recommendation Engine
The recommendation engine was validated through three parallel streams: Mac Experts reviewed accuracy; User Testing IDIs assessed desirability, usability, and content models; and machine testing validated product distribution across all possible Q&A configurations, ensuring that no combination produced a wrong or misleading recommendation.
Press & Public Response
The launch generated coverage in Forbes, TechRadar, and Techloy, with reviewers praising the tool's brevity, plain language, and intelligent recommendations. Social response on X and MacRumors reflected genuine user satisfaction - and calls for Apple to expand the tool to iPhones, iPads, and Apple Watch.
Dyson x ELVIS London
Dyson Care
Consumer Electronics Voice AI IoT Device Experience Award Winning

We developed Dyson Care, an Alexa Voice Assistant that helps Dyson cordless vacuum owners troubleshoot problems and maintain their products through natural conversation. Designed to deflect customer service volume while enhancing the support experience, Dyson Care provides 24/7 assistance and positions Dyson as the first home appliance brand to offer conversational customer service. The project won Best Use of Audio at the 2023 Digital Impact Awards.

Voice-First Troubleshooting & Diagnostic Flows
We designed multi-turn diagnostic conversations that guide users through problem-solving via natural dialogue. Users describe an issue, and Dyson Care asks clarifying questions, provides step-by-step troubleshooting, and delivers solutions conversationally, understanding intent, accessing product knowledge, and translating technical information into accessible guidance.
Proactive Maintenance & Reminder Systems
Dyson Care enables users to set maintenance reminders and receive proactive performance tips, integrating Amazon's send-to-phone and Set Reminder features to bridge voice interactions with ongoing engagement. We identified this as a key differentiator, with none of Dyson's competitors offering similar levels of support at the time of launch.
Multilingual Deployment & Scalability
We launched Dyson Care across both the US and German markets, handling localisation of conversational flows, voice personas, and technical terminology. The solution architecture was designed for scalability, enabling rapid deployment while maintaining consistent quality across markets.