Author transcript
Rhys Fisher: Hello and welcome to CX Today. I’m Rhys Fisher, Associate Editor, and today I’m delighted to be joined by Frédéric Durand, the founder and CEO of Diabolocom. Frédéric, thank you for joining me. How are you doing today?
Frédéric Durand: Good afternoon, Rhys. A pleasure to meet you again. Doing well.
Rhys Fisher: It’s been a little while since we had a chat, so I’m looking forward to getting into things again. We’re discussing what I think is a really interesting topic today, based around some comments you’ve made recently which, I think it’s fair to say, are pretty bold and also quite refreshing, around AI’s position right now in CX and the contact center space, and perhaps what you think AI’s position should be moving forward. It’s based on some really interesting points you wrote in a manifesto recently for Diabolocom, so I’d recommend everyone check that out.
I wanted to start with one of the main points raised in that manifesto, where you take direct aim at what you term the “empty contact center” vision that’s being sold by certain parts of the industry right now. As a technology vendor yourself who is investing in AI, what are you seeing in the market that makes you believe the empty contact center, or full autonomy, is the wrong goal?
Frédéric Durand: Thank you for the question, it’s very interesting. There’s been much talk in the market about that. According to us, there is a real gap between the AI story that is being sold in the industry and what actually happens on the contact center floor. Usually demos are perfect and seamless, but in production it’s a totally different story. Agents are still switching between systems, they are still searching for context, and they are dealing with the same operational friction all the time.
So the technology itself is no longer the main problem. The problem is how it is packaged and deployed. Too many projects start with “where can we put AI?” instead of “what problem are we trying to solve for the customer or the agent?”, which is the question that should be asked. This often leads to generic technology being forced into workflows it was not designed for.
The gap between ambition and operational reality is exactly what we wanted to test in our recent US research with CCW Digital, and the findings reinforce what we were seeing: 89% of leaders still expect humans to play a key role in customer support. This is why we challenge the empty contact center vision that we mention in our manifesto. We believe the point of AI is not to remove people. It is to remove what stands between people and the customer, and ultimately help humans become more exceptional at the things only they can do.
Rhys Fisher: Thanks for that. Like I said at the start, these are really refreshing remarks coming from someone in your position, running a vendor that supplies this technology to the contact center, to hear that you’re still so focused on the human element. I thought an interesting part was where you used the word “presence” in the manifesto, presenting it as an alternative to automation. But presence isn’t a product feature; it’s not something you can put on a spec sheet. So what does it actually mean in operational terms for the contact centers you’re working with?
Frédéric Durand: Yes, presence is definitely not a feature. Presence is not speed or availability. Presence is being there when it matters. Operationally, presence means removing the distance between the agent and the customer, and we think about it in three ways.
First, it means being present in the critical moment. The agent should have the right context and answer in front of them, instead of searching across multiple systems while the customer waits. Waiting time for customers has always been a nightmare in the contact center.
Second, presence means being present wherever people cannot be. That means across channels and outside business hours, for instance. During those moments, the conversation should carry forward, so the customer never has to start again.
Third, presence is also being present in your data. Customer data should remain safe, because data is absolutely critical in the era of AI, and the intelligence should work where the data lives rather than requiring you to send it elsewhere.
So in these examples, presence is the wanted outcome, and automation is one of the tools that can help create it. This is what presence means to us at Diabolocom: using technology to make the company more available, more informed and more useful for the end customer.
Rhys Fisher: Thanks, Frédéric. I really like that differentiation between AI’s role and the human role in the contact center. I wanted to play devil’s advocate a little here. A lot of vendors talk about AI that augments rather than replaces humans; it’s quite a common phrase now. But many would argue that vendors are guilty of being all talk and less action when it comes to this in practice. So what is the meaningful difference between saying it and actually building for it and delivering it?
Frédéric Durand: Thank you for the question, Rhys. The difference is whether there is an actual positive change in the agent’s day: fewer searches, fewer clicks, less after-call work, better context. If AI simply adds another tab, an alert or a dashboard, that isn’t augmentation.
At Diabolocom, we also make a distinction between tasks and the job itself. AI is very good at tasks; everyone knows that. But in customer service, the real job is the relationship: understanding someone, exercising judgment and making things right. AI should clear away the routine so people have more capacity for that work. This is where their value is.
That’s why at Diabolocom we believe AI for customer service has to be purpose-built around the realities of the contact center. Voice, routing, customer history, knowledge, compliance and integration all have to work together. You cannot simply insert a generic AI model into the workflow and expect the employee to adapt around it. That doesn’t work at all.
And finally, “people-centered” has to mean something in practice. The research I mentioned showed that only 34% of organizations measure employee sentiment about AI’s impact on workflow, and only 39% actively gather direct employee feedback. So there is great room for improvement there.
Rhys Fisher: Absolutely. And again, I like the choice of language you’re deploying there. That segues nicely into something else I wanted to ask you about from the manifesto, where you talk about the “superhuman operator”: the agent who, with AI handling the volume and the tasks, can focus almost entirely on judgment and the customer in front of them. How far away are we from that being the norm? And if we’re not there yet, what is standing in the way?
Frédéric Durand: We all see the major technological progress, so I would say that on technology we are quite close. But from an organizational point of view, we are much further away.
Automating a task today is relatively easy, as I already mentioned. Redesigning roles, retraining people and integrating AI into the way a team actually works is not that simple. The superhuman operator is definitely not a simpler version of today’s agent. It’s a totally different role. The human spends less time searching and processing, and more time on judgment, emotion, exceptions and complex customer situations.
The issue is that the technology is moving faster than the workforce model around it. Again, our research found that only 17% of leaders believe their teams are ready for next-generation work. So there is a lot of training to do. This is as much a management challenge as a technology challenge. The technology will not do the job by itself.
Getting there means owning the transformation, not just deploying the technology. It requires retraining people, redefining roles and responsibilities, adapting management, and making sure employees are prepared for a job that will inevitably become more complex. This is the main point.
Rhys Fisher: I think that’s a really important point. Any of us in the tech space are a little bit guilty of this: we like to see the new tool, the new feature with the bells and whistles, the AI that’s automating things and doing live translation and whatever else. Perhaps we overlook the organizational element you’ve touched on there.
Finally, Frédéric, if a CX leader is watching this and they’ve already bought in and invested in the empty or autonomous contact center vision, what would your honest advice be to that person?
Frédéric Durand: I definitely wouldn’t tell them to abandon automation and their investment. I would tell them to reconsider the finish line, what they want to achieve. The goal shouldn’t be autonomy for its own sake. It should be a better customer relationship, because this is what matters for their customers and for the good of their company. AI is a means to get there. It’s not the CX strategy itself.
I would also tell them to look at what the technology has actually changed in production. Has it removed customer friction? Has it made agents more capable? Has it improved the moments that really matter?
And then I would tell them to look at what AI leaves behind for people. As routine work disappears, human interaction becomes more complex and demanding. So agents need the context, the knowledge, the training and the authority to handle that shift.
Rhys Fisher: Thanks, Frédéric, that’s a really great chat. Like I said early on, it’s so nice to hear candid, honest conversations from people in your position about AI’s current role. We all know it has incredible potential, but it’s good to hear people talk about where it can fall short and what we need to do to really maximize that potential. I think the audience is going to get a great deal from that. Thank you very much for your time.
Frédéric Durand: Thank you very much for your time, Rhys. It was a pleasure.
Rhys Fisher: And I also want to quickly thank our audience for tuning in. If you enjoyed this, please do remember to like and subscribe to the channel, and head over to cxtoday.com for more stories like these. Until next time, thanks for watching.
