AI in Contact Centres: the future of customer experience

Introduction

Craig Farley:
Okay. Hi. Welcome to the IPI Converse Podcast. This is my first time on the podcast. I’m Craig Farley, at Solution Consulting, IPI.

Iain Benett: It’s also my first time in the IPI Converse suite, so I’m looking forward to it. Thank you for having me.

Craig Farley
You’re welcome. And we’re going to be talking about AI today. So you’re a bit of a specialist in AI?

Iain Bennett:
That’s what they say, yes. I am an AI specialist. I try to keep up to date with everything, and it’s moving very, very fast, as you know. But yes, they cast me as an AI specialist, so let’s see how that pans out.

Perspectives on AI

Craig Farley:
So let’s start on a broad topic. High level, what’s your perspective on AI in general? What do you like about it?

Iain Bennett:
For me, I’m split down the middle on this one. A lot of people probably don’t know it because I cover it up quite a bit.

I have two sides of AI. From a techie point of view, I absolutely love it - the advancements, playing with it, what I can do with it. It’s amazing. The things we’re seeing are mind-blowing.

But then there’s the other part of me: I look at my sons and think about the future of jobs, and that worries me. We don’t know what the future really looks like. There are different models, but nobody can say for sure. That’s the worrying part - that we simply don’t know.

That said, it’s absolutely amazing what AI can do. In areas like medicine, the future could be brilliant. But there’s always a scary part as well.

Craig Farley:
Yes, similarly, I’ve got two kids. It’s hard to know how to advise them career-wise. It’s too early to know how it’s going to pan out. I saw something similar with RPA years back. Bots were hyped as if they’d take over everything, but it didn’t play out that way. We’re probably still in the hype phase with AI, but there’s definitely more impact to come.

Iain Bennett:
Absolutely. AI will impact every single industry. That’s the amazing thing - it won’t leave any sector untouched.

One of the best things is experimenting with it. My sons love the fact that I train AI models and get them involved. Although, like your kids, they don’t take advice from me - they’ll listen to someone else saying the exact same thing instead!

Craig Farley:
That sounds familiar. I also see a lot of potential for AI to augment people, allowing them and society to do more. But I worry about the downsides too - cost-cutting in creative industries, homogenisation, and not letting people flourish at what they’re good at.

Iain Bennett:
Exactly. Just yesterday the government announced new initiatives around AI. That shows how far-reaching it is. I hope the effect will be positive, but no one will escape its impact.

AI in Contact Centres

Craig Farley:
So let’s move into the contact centre space. How is AI transforming contact centres, and what do you see for the future?

Iain Bennett:
It’s transforming them massively.

AI routing is a huge step forward. In the past, bots relied on specific keywords. If they didn’t get them, customers had to repeat themselves until frustration built up. Now, with generative AI, systems can understand natural speech, figure out what the customer wants, and route them appropriately.

That could mean self-service, or it could mean sending the customer to the right agent straight away. From that point of view, AI is massively changing contact centres.

For agents, features like auto-summarisation and auto-wrap are really coming into their own. They reduce handling times and admin work, cutting call times significantly.

Looking ahead, translation will be huge. Soon it won’t matter what language you speak - you’ll be able to converse naturally, and AI will translate in real time. For example, a customer could speak German, AI would translate to English for the agent, then translate the agent’s reply back into German. It’s incredible - a real “Star Trek moment.”

Craig Farley:
Exactly. We used to call it “conversational AI,” but it never felt conversational - more stilted, rigid. Now with generative AI voice bots and assistants, it’s far more natural.

Iain Bennett:
Yes, and training conversational AI has changed too. Before, we had to script every possible way a customer might phrase something. Even then, customers always found new ways to say things.

Now, with large language models, you can just ask the AI to detect frustration or vulnerability without having to list every phrase. It understands, thanks to its training. That reduces frustration for customers across age groups, since it can interpret different ways of speaking.

And the voices are so much more natural now. In the past, we hired voice artists to record prompts, which was slow and costly. Today, AI-generated voices are fast, flexible, and realistic.

But that brings us to ethics: do we tell customers when they’re speaking to AI? Some companies announce it, others don’t. Personally, I think we should.

Craig Farley:
Absolutely. Anything generated by AI - text, speech, images - should be labelled as such. Some countries, like Spain, are already moving toward laws requiring this.

Being clear avoids frustration. Also, AI doesn’t have empathy - at least not yet. If a customer thinks they’re talking to a person but feels no empathy, it could damage the brand. Transparency helps.

Iain Bennett:
Agreed. Empathy and even sarcasm detection are coming, but for now, AI doesn’t fully have those capabilities. It’s moving fast, but we must be honest about what it can and can’t do.

Innovative AI Use Cases

Craig Farley:
What are some of the most innovative ways that AI is enhancing customer interactions, and how does this improve customer service within contact centres?

Iain Bennett:
We’ve spoken about AI routing, but from the agent’s perspective, auto-summarisation and note-taking are game changers.

When I used to be an agent, my notes were terrible! Having something that listens in and writes structured notes for me would have been invaluable. It also helps with disposition codes and after-call work.

Real-time transcription is another big one. It can surface knowledge articles or compliance information as you speak, like having a personal assistant feeding you answers.

This saves time, improves accuracy, and boosts first-contact resolution.

Craig Farley:
When I was an agent, my “assistant” was Clippy - or more often, a desk carousel with 100 pages of phone numbers, knowledge articles, and compliance statements. We had to memorise huge amounts.

Now agents don’t need that depth of industry knowledge. You hire for empathy and communication skills, not memory. AI provides the knowledge.

Iain Bennett:
Exactly. You can be a brilliant agent in any industry, because AI supplies the domain expertise in real time.

Benefits for Businesses

Craig Farley:
We’ve started to touch upon some of the benefits. What are the most noticeable improvements businesses might see in terms of efficiency and customer satisfaction when they implement AI?

Iain Bennett:
Efficiency is key. AI helps agents resolve issues faster and improves first-contact resolution - a huge metric in contact centres.

It also reduces admin through auto-summarisation and enables self-service. That frees agents to focus on complex, empathy-driven tasks.

For businesses struggling with long queues and high wait times - especially post-COVID - AI can deflect simple queries and speed up routing. That cuts frustration and increases customer satisfaction.

Craig Farley:
Yes, old IVR menus with “press one, press two” drove people mad. AI routing evolves that. Customers just explain why they’re calling, and AI sends them to the right place -  whether that’s self-service or the right agent.

Iain Bennett:
Exactly. It eliminates unnecessary transfers, avoids repeating yourself, and builds confidence.

Challenges and Misconceptions

Craig Farley:
Of course, with every new innovation comes challenges. What steps can businesses take to overcome these and get the most from AI adoption?

Iain Bennett:
The biggest one is choosing the right partner.

There are a lot of so-called “AI experts” out there. You need someone who understands both AI and your business, and who can advise honestly - sometimes even saying you don’t need AI.

The rush to adopt can cause problems: poor implementations, misuse, or brand damage if prompts are exploited.

Focus on real business problems. If AI can solve them, brilliant. If not, fix them another way.

Craig Farley:
Exactly. Not everything needs AI. Some issues can be solved with integrations or process changes.

And businesses need people to “own” AI internally. You can’t just turn it on and expect miracles. Someone has to monitor, tune, and optimise it.

Iain Bennett:
Yes. AI is easier to deploy than older tech, but it still needs skills. It’s not a silver bullet.

We’re moving into agentic AI - systems with sub-level thinking that can plan and act. That’s the next big leap.

Dispelling the Myths

Craig Farley:
There are also misconceptions about AI. What can businesses and partners do to dispel the myths?

Iain Bennett:
First: AI is not a silver bullet.

Second: the term “AI” is overused. Everything gets branded AI - from phones to kettles! But AI isn’t one thing; it’s a toolbox.

People’s experiences differ. I use AI daily - others barely touch it. That shapes perceptions.

And we need to move beyond Hollywood fears. AI isn’t self-aware. Yes, it’s advancing fast - in self-driving, robotics, and more - but it’s not the Terminator.

Craig Farley:
Yes, in a few years we may see affordable humanoid robots in homes. That raises big questions about how comfortable we are with them.

Iain Bennett:
Exactly. The term “AI” doesn’t do justice to its complexity. It’s almost like dealing with a new species - an entity that learns and adapts.

AI Agents and the Future

Craig Farley:
Let’s end with the elephant in the room: AI agents. Where do you see that taking us?

Iain Bennett:
In some ways, yes - they’ll be powerful and autonomous. But most of what’s called “AI agents” today isn’t truly agentic.

A real AI agent has sub-level thinking. For example, if I tell my AI assistant “Mike” to organise a business trip to New York, it won’t just book a flight. It will check my calendar, company travel policies, preferred hotels, and arrange transport.

That’s sub-level thinking - planning multiple steps autonomously.

In future, AI agents will interact with humans and with each other. Imagine my agent booking travel by negotiating directly with an airline’s AI agent.

Technologically, this means moving beyond APIs to new frameworks like MSPs, enabling agents to communicate more flexibly.

Craig Farley:
Yes, the AI side - understanding and conversation—is already advanced. The challenge is integration: connecting all the backend systems seamlessly.

Iain Bennett:
Exactly. That’s where we’re heading - towards seamless agent-to-agent integrations.

For consumers, this will feel like having a personal assistant. You’ll simply say, “Do this,” and it will handle everything.

For businesses, preparation is key. Data readiness is critical - AI runs on data. Even if you’re not using AI yet, clean and centralise your knowledge bases, integrate systems, and define AI guidelines.

And remember: once AI learns something, it can’t “unlearn” it. Be careful with sensitive data. Always ask vendors where your data goes, how long it’s stored, and who has access.

Key Takeaways

Craig Farley:
That’s a good place to wrap up. To summarise:

- AI brings huge benefits, especially in efficiency and customer experience.

- It’s not a silver bullet - you need strategy, guidelines, and preparation.

- Do the groundwork now: clean data, integrate systems, build knowledge bases.

- Don’t rush into AI for the sake of it, but don’t ignore it either. Start learning today.

Iain Bennett:
Absolutely. The best time to start learning AI was yesterday. The second-best time is today. AI changes daily - sometimes drastically. So businesses and individuals need to keep adapting.

Craig Farley:
Exactly. Thanks for the conversation - it’s been a really good talk.

Iain Bennett:
Thank you.