top of page

Modernising Customer Service with AI: How Conversational AI can help businesses, help themselves.

Updated: Jun 8

Conversational AI is revolutionising the way that businesses interact with their customers.

Coversational AI man frustrated with customer support.

Whilst these dialogues may look different from traditional face-to-face communication, they offer a new level of convenience and efficiency.

The ability to analyse large amounts of data, continuously learn and improve, whilst adapting to the preferences of the individual is what makes conversational AI such a hot topic.

As I converse with these virtual agents, rather than their human counterparts, I find myself analysing the similarities and differences in our exchanges, the pros, cons, and pitfalls of each. But I always ask the same question, what makes a truly great customer service conversation?

What are the similarities between AI and real life…

For instance, let’s consider one of the most common issues handled: refunds.

In the Store…

A customer walks into a store, goes up to the shop assistant and says, ‘Excuse me, I need to return this item.’ They explain the issue, a couple of follow up questions, and then the assistant processes the return for a refund. The customer receives an apology, attains their refund or exchange, says thanks and leaves the store.


A customer initiates a conversation with a conversational AI. The customer says, “Can I get a refund”. The AI will ask a series of questions to determine the reason for the refund request, and if appropriate, provide a refund automatically; or by directing the customer to a customer service representative. The AI apologises for the inconvenience, the customer gets their refund and gets on with their day.

Whilst the channels of communication differ, both conversations involve the same components:

  1. The customer is dissatisfied and needs a resolution.

  2. The assistant, human or AI, is working towards delivering a positive customer experience, in this case – processing a refund.

  3. Provided, criteria are met. The customer received the correct outcome for their request.

What happens when it’s not quite that simple…

Take for example the following scenario between a customer and a human shop assistant:

Customer: Hi, I’d like to request a refund for my order I bought online a couple of weeks ago.

Shop Assistant: Hi there, I’m sorry, but refunds are not available for this item unfortunately.

Customer: Really, why? When I received it, it was damaged.

Shop Assistant: I understand, and I’m sorry about that, however our policy states we cannot process refunds for damaged items more than 5 business days after the purchase is made.

Customer: This is ridiculous! How can you not accept returns for damaged items?

Shop Assistant: I apologise for any inconvenience I have caused, unfortunately this is our policy.

Customer: Fine – I’d like you to go and get your manager.

Shop Assistant: I’m sorry, but my manager isn’t available currently. Is there anything else I can assist you with today?

Customer: That’s just great. Good day, I will not be buying from you again.

Shop Assistant: My apologies once again. Thanks for chatting to me today –if possible, please could you leave me some feedback on our conversation?

Customer: Useless.

Familiar… I thought so.

If we replace the ‘Shop Assistant’ with ‘Virtual Assistant’, you can see how this conversation could quite easily take place online.

What’s most interesting to me is that the shop assistant in this case technically did nothing wrong – it’s emphatic, highly responsive, acknowledges the customer requests accurately and operates 100% in line with the business processes outlined for them.

This time, with a virtual assistant, but the conversation takes place like this…

Customer: Hi, I’d like to request a refund for my order.

Virtual Assistant: I’m sorry to hear that, can you please provide me with some more information about the reason for your refund request?

Customer: Sure, the product I received was damaged when I received it.

Virtual Assistant: Thank you for letting me know, when did you purchase this order?

Customer: A couple of weeks ago.

Virtual Assistant: I apologise for any inconvenience this may cause, however I’m unable to process your refund for you right now, as it’s outside of our policy rules. However, I am going to escalate this to my manager who will be in touch shortly, I’m sure they can help.

Please could you leave me some details?

Customer: Thank you, I really appreciate that. Here you go…

Virtual Assistant: Thank you, I’ve passed these details on via email. Is there anything else I can do for you?

Customer: No, that’s all, thanks again.

Virtual Assistant: No problem, please could you leave me some feedback so I can continue to learn and improve?

Customer: Excellent service, will be shopping with you again.

Both conversations, could easily have taken place with the same business and with either an AI or a human agent…

The differences lie in what process the assistant is following.

So, are conversations really changing, or do business just need to keep up?

At the core of any effective conversation is a goal.

Ami is a goal-oriented conversational AI, that operates using your business knowledge.

Ami uses conversations to work with your business on shared goals.

But how do you know which processes are really affecting your bottom line…

Ami can help…

Ami identifies where knowledge gaps or unhelpful processes, like the refund policy in our conversation above, exist within your business.

Interactions like the above can be flagged by Ami and highlight underlying issues causing customer frustration, to enable a business to take steps to address them.

These insights are invaluable and enable businesses to:

  • Keep their knowledge base up to date.

  • Modernise their processes, especially the frustrating ones.

  • Provide additional training for human assistants handling the same issues.

Conversations are the informant & the resolution.

It’s not the conversations that are changing, but our expectations of the outcome. We want the same brilliant conversations from an AI and a human alike, and we want them quickly with no friction.

Without effective processes in place, an AI may struggle to provide accurate, honest, and helpful responses to your customers, like the shop assistant in our first conversation, which could translate into negative reviews and losing business.

Conversational AI is a game changer. Ami works with your organisation to increase automation through continuously improving conversations. This frees up human customer service assistants, to handle more complex issues and provides a platform to deliver more personalised support to customers who truly need it. It improves efficiency, running costs all whilst helping you define and refine better business processes.

It’s vital that effective processes underpin your business to equip your AI with everything it needs take care of your customers, but let Ami tell you where to start. Schedule a quick 15-minute demo.

API & Webhook Support

For further information on Ami's API's and Webhooks please contact the support team using the button below and selecting API Information. 

bottom of page