How Prudential is shaping user-centred design in the age of generative AI

Picture of Kane Simms
Kane Simms

Conversational AI has moved fast over the last few years and changed how we design for user experience (UX). On the VUX World podcast, Alex Shin, Senior Product Designer at Prudential Financial, talked about the journey from the early days of natural language understanding (NLU) to the latest wave of generative AI.

From wordsmiths to experience custodians

In the old days, conversation designers crafted every word, punctuation mark and dialogue flow. This was key in the NLU era, where pre-scripted responses had to be carefully planned and mapped out to avoid confusion and ensure user satisfaction. As Alex said, “Words were my medium”, highlighting the level of detail required to get the user the right and helpful response.

But with generative AI, the landscape has changed. Designers now have to maintain control of the user experience when the AI generates the response. Alex said that with generative models, some outputs can’t be pre-written or tweaked in advance. So we have to adapt from controlling every word to overseeing the whole experience: making sure the response is relevant, understandable and aligned to user expectations.

Human-centred design in a generative world

At Prudential, the design team has been focusing on user-centred design principles. Rather than just relying on technical performance metrics like confidence scores, they prioritised whether the response met the user’s needs. Alex said success isn’t just about speed or accuracy, but whether the AI actually helps the user achieve their goal.

One of the strategies was to bring the design and technical teams together. Alex highlighted the importance of cross-functional teamwork where designers, machine learning engineers and developers contribute to refining the AI’s output. This ensures the generative model aligns with design expectations – for example, breaking up complex responses into bite-sized chunks, using the right tone and clarity.

Also, Alex’s team at Prudential built in flexibility into their offerings. They give users choices, they can switch between AI chatbots, live chat with human agents and even voice options. This acknowledges that different users have different preferences – some value the speed of AI, others want the reassurance of a human touch.

A more empowered design role

Generative AI hasn’t reduced the importance of design – it’s increased it. Designers are now ‘experience custodians’. Instead of just crafting dialogue, designers now ensure the AI response meets the highest standards of usability and user satisfaction. This means overseeing everything from tone and structure to the overall integration of the AI within the platform.

At Prudential, this user-centred approach has paid off. Alex said their customer satisfaction scores are above industry average, with users scoring an average of four out of five. This is from balancing technical innovation with human needs, so the technology works and serves the customer.

Designing for the future

Alex Shin’s journey at Prudential shows how conversational AI design has moved from writing lines of dialogue to orchestrating seamless experiences. UX design in the generative AI era is less about micro-managing words and more about shaping the whole user journey. This requires cross-disciplinary collaboration, deep understanding of user needs and continuous improvement.

For other designers looking to get into generative AI, Alex’s advice is simple yet powerful: focus on the human at the centre of every interaction. Use every experience as an opportunity to learn, reflect and apply those learnings beyond one project. In this way, designers become not just content creators but experience custodians.

In an era where generative AI is changing how users interact with technology, Prudential’s approach is the blueprint for success: combine technical innovation with human-centred design to build experiences that resonate.

Share

A customer's AI agent just called. Now what?

This research looks at how businesses are responding to personal AI agents.

Take the 5-minute survey

Everyone who takes part gets first access to the results

Weekly newsletter

The latest in AI-powered customer experience. Make better strategic decisions with the help of our weekly newsletter.

Kane Simms

A strategic AI advisor who, for the past decade, has helped business leaders and product owners transform customer experience using conversational and generative AI.

Share

✓   Link copied

A customer's AI agent just called. Now what?

This research looks at how businesses are responding to personal AI agents.

Take the 5-minute survey

Everyone who takes part gets first access to the results

Weekly newsletter

The latest in AI-powered customer experience. Make better strategic decisions with the help of our weekly newsletter.

Top articles and podcasts

Related content

Why voice AI agents pass in the lab and fail on real calls: Liam Dunne (Klearcom)
Liam Dunne, Co-founder and CEO of Klearcom, joins me to talk about why voice AI agents can pass every lab test and still fail on real calls, and why testing the whole journey is becoming mandatory for global deployments.
Brands are split on personal AI agents. Where does yours stand?
Personal AI agents can now phone businesses and pay on a customer's behalf, but some brands are already blocking them. I'm running research to find out where businesses stand.
AI is stripping brand value: Matt Clare (UJET)
Matt Clare, VP Product Marketing at UJET, joins me to explore what happens when your customer's AI agent is the one calling, and why AI is stripping away the brand and experience value businesses have relied on to stand out.
How Sling TV’s AI agent converts 25% of sales conversations
Sling TV's AI agent converts nearly 25% of sales conversations and the team behind it shared how it grew from answering simple questions to selling at scale.

AI Breakfast: Keeping Control of Enterprise AI Agents. London. 26th November.

Meet VUX at CCW Europe, Amsterdam, October 5-7

Register now: Why your contact centre playbook is obsolete