After 12 years of operation in Dubai, ZIWO has evolved from a voice-centric contact center solution to an AI-powered customer experience platform.
In this interview, CEO and Co-founder Renaud de Gonfreville shares insights into how the company is revolutionizing customer service across emerging markets.
ZIWO's cloud-based solution handles everything from voice calls to WhatsApp messages, with a crucial focus on sovereign cloud deployment to ensure data compliance. The platform's integration of AI has transformed how businesses handle customer interactions, automating routine tasks while enhancing agent effectiveness.
"We don't just do AI just for the sake of doing AI," de Gonfreville emphasizes. "We first analyze where it can add value." This approach, he says, has led to innovations in handling complex customer interactions, from identifying at-risk customers to providing agents with real-time insights during calls.
With 130 employees and a growing presence across the GCC, Iraq, Egypt, Morocco, and expanding into African markets like Kenya and South Africa, ZIWO is particularly focused on regions where voice communication remains crucial. The firm’s partnership with Huawei, for instance, has enabled advanced AI capabilities, including GPU-powered solutions in Saudi Arabia.
De Gonfreville also notes an interesting observation about AI interactions during our conversation: "People are always very polite when they're asking a question to ChatGPT, which is not always the case when they talk to human people." This insight reflects the evolving nature of customer service expectations, which he compares to the Netflix effect on digital service standards.
Looking ahead, de Gonfreville sees customer experience becoming an increasingly critical differentiator for brands, especially in highly regulated industries like banking where product differentiation is limited.
Want to hear more from de Gonfreville? Tune in now!
Steve Saunders (00:15):
Renaud, tell me a little bit about ZIWO.
Renaud de Gonfreville (00:16):
So, it's very simple. ZIWO is a software. It's a software as a service. We've been established in the region of Dubai for the past 12 years.
Steve Saunders (00:26):
12 years?
Renaud de Gonfreville (00:27):
Yeah. So, the software is providing the companies with a solution to power their call center and their contact center. So if you are a company, for instance, and you receive a lot of calls or a lot of messages in your customer care, for instance, people are angry or they want to ask questions about your product, then we are the software that take all the calls and dispatch them to the right people.
Steve Saunders (00:52):
Oh, okay. So, you are like a hub which directs the ...
Renaud de Gonfreville (00:56):
That's exactly right. Yeah, and will-
Steve Saunders (00:57):
Yeah, and channel the anger to the right person? Yes, okay.
Renaud de Gonfreville (00:57):
Yeah, exactly. Yeah.
Steve Saunders (01:01):
That's a lot of responsibility.
Renaud de Gonfreville (01:03):
Yes, it is.
Steve Saunders (01:03):
But not just for voice, not just for voice?
Renaud de Gonfreville (01:05):
So initially, it was for the voice, which is the most complex, because when people are angry, they want to talk to someone. So, the voice is very specific in our business, and that's a big differentiator that we've got, but now it's all the channels, it's WhatsApp, emails. We plug in on top of the AI now so that we can automate a lot of the transactions and help the people that are supposed to respond to get valuable information and insights during the call so that they are more effective.
Steve Saunders (01:36):
So, it's Cloud based?
Renaud de Gonfreville (01:38):
Its Cloud based, but a local Cloud. That's very important. When it comes to the voice, you want the voice to stay in the country, that's the legal part of it.
Steve Saunders (01:50):
So, this has to do with sovereign Cloud?
Renaud de Gonfreville (01:52):
Exactly, yeah, sovereign or in-country.
Steve Saunders (01:55):
Yeah. And what is the role of AI now, because we hear about AI a lot?
Renaud de Gonfreville (02:00):
Really, you do.
Steve Saunders (02:01):
But how do you use it? Once or twice just this year.
Renaud de Gonfreville (02:04):
Yeah. No, the way we see the technology is just a means of solving a problem for our customers. That's very important. We don't just do AI just for the sake of doing AI. So, we first analyze where it can add value. And clearly, in the call center industry, there are a few challenges that people have. First, it's finding the right people, because it's not that easy. I mean, you've got a very high attrition in that industry. And one of the reasons is that a lot of the transactions that people have to handle, they are very mundane, they are not super interested, so the AI is a way to automate all this basic stuff that people don't want to do. And at the same time, it's a way to ensure that the quality and the consistency of the answers is always what the company wants. So, it's a way to do quality insurance, and also, it's a way to provide, very quickly, complex information to the person in charge of answering the questions.
Steve Saunders (03:08):
So, they're prepared?
Renaud de Gonfreville (03:09):
Yeah. So that in the AI, you can, for instance, spot very quickly if a customer is just talking about quality issues or price issues, naming a competitor, and then you can add some flags to the conversation. Maybe this customer is at risk, and you want to trigger some automation behind that, maybe to forward the discussion to a product manager or to a manager of the company. All these things become available thanks to the AI.
Steve Saunders (03:39):
So, you've been around for 12 years, so at some point you didn't have AI in the system. I mean, what is the customer service experience like with AI now?
Renaud de Gonfreville (03:47):
I mean, surprisingly, the people are always very polite when they're asking a question to ChatGPT, which is not always the case when they talk to human people, because-
Steve Saunders (03:56):
Yeah, they're worried they won't agree with them.
Renaud de Gonfreville (03:57):
We don't know what, yeah, yeah, exactly. They're a bit scared about the AI in the future. Well, I mean, the people normally are calm and polite. You've got some extreme reaction of course. But no, interestingly, in customer experience, the people have different expectations, and the expectations are going up and up and up, because you've got this large player. I would just take the example of Netflix. They have completely transformed the customer experience and the way the people are experiencing the Cloud and the internet. Before Netflix, you were able to wait for minutes, even hours, to download the video. Now, it needs to start instantly. So, in customer experience, the expectation of the customers is always going up and up and up. So if you're not investing, as a company, into new technologies for powering your customer experience, then you are missing something.
Steve Saunders (04:52):
Absolutely. Yeah. How big is your company now?
Renaud de Gonfreville (04:55):
So now we're 130 people. We are growing very fast, and we're expanding in the region. So, our strategy is two-fold. It's to invest in the product, to provide more and more products, and to grow the average basket of our customers and to open new countries. And as a company, we've got a strategy to expand into emerging countries, so we are now covering entirely the GCC countries, Iraq, Egypt. Now we just opened Morocco, and we are going down to some of the economies in Africa like Kenya, South Africa.
Steve Saunders (05:33):
Okay. So, you're focused on the global south now?
Renaud de Gonfreville (05:36):
Yeah, the emerging countries, but not because there are emerging countries, but because they all have in common these specificities around the voice, which we handle very, very well. So that's where we believe that we have leverage in this.
Steve Saunders (05:53):
Are you working with Huawei?
Renaud de Gonfreville (05:55):
Of course. What we like about Huawei is that they have a bit of the same strategy. I mean, of course, they are very strong on the AI, so they provide compute in countries where we're confined, yeah, recently where these big customers in Saudi that where we are launching a new AI concept. And we needed some GPU, you probably know what GPU is?
Steve Saunders (06:21):
Yeah. Yes.
Renaud de Gonfreville (06:22):
You need that for complex AI calculation, and they were the first one to provide that in Saudi, so we started with them on that project.
Steve Saunders (06:33):
Yeah. Fantastic. And do you have any predictions about where the industry is going to go, where your industry segment is going to go?
Renaud de Gonfreville (06:43):
Our industry?
Steve Saunders (06:44):
And what do you think is going to happen?
Renaud de Gonfreville (06:46):
Yeah. In the customer experience, I mean, the expectation of the people is higher and higher. And what we see is that customer experience is a big, big differentiator for many brands where they can't really differentiate on the product side. Think about the bank, for instance. It's so regulated that the only way you can make a difference is through the customer experience. Also, if you think about a company like Starbucks, what's the reason you would go to Starbucks to pay your latte 25 dirhams, which is what, the equivalent of $8? Why would you do that? I mean, it's just a cafe latte. And the reason is because you've got this nice barista picking the cup, asking you questions, putting handwritten names, and that's customer experience. So, with all these experiences, people are ready to pay for that. So, on one side it's vital, if you don't do it properly, you will lose customers, they will trash you on social media. And on the other side, if you do it properly, I mean, they will come to you, and they will be ready to pay extra money just for the experience.
Steve Saunders (07:54):
Oh, absolutely. No, no, we see it all the time. I think internet technology and AI, it creates accountability.
Renaud de Gonfreville (07:59):
Yeah.
Steve Saunders (07:59):
The AI that you use, the last question, you trained it or is it general purpose or how does it work?
Renaud de Gonfreville (08:05):
No, we train it big time, because you can simply ask, okay, ChatGPT, you take it and it speaks 150 languages, answer a lot of questions, but in the context of a company, you need to train it to your particular context. In a call center, you've got a lot of background noise. So, if you give that directly to ChatGPT, it will pick some background noise and think it's part of the conversation. So definitely, you need to train the models. So, we have LLMs that we have selected, and we train them in the context of our customers in two main dimensions. First, is the context of the company, so to plug-
Steve Saunders (08:51):
A good...
Renaud de Gonfreville (08:52):
Yeah, to plug their own data. I mean, think about a restaurant. You need to tell them what's the menu, what's the pricing, what are the promotions, so you need that particular training. And also in the region, we've got a lot of different dialects and specificities, so we need to train. Talking about Saudi Arabia, we trained our LLM with real Saudi people to correct and to try and, so this fine-tuning exercise is extremely important to bring from a 50% accuracy to close to 100%, and then it really starts to create value to our customers.
Steve Saunders (09:30):
Merci, Monsieur.
Renaud de Gonfreville (09:30):
Merci beaucoup. C'était un plaisir.
Steve Saunders (09:30):
Oui.
Renaud de Gonfreville (09:30):
Merci.