
A Voicebot, or voice assistant, is a conversational robot capable of interacting vocally with a user through speech. In other words, a Voicebot is a virtual agent capable of understanding, interpreting, and analyzing requests expressed by a human in order to respond vocally in natural language, that is to say, in the same way a human being would speak.
A Voicebot is built using artificial intelligence, bringing together all the technologies that enable the robot to learn how to understand human language and the intentions conveyed through spoken words. When we talk about intelligence, we often think about the brain. The robot’s “brain” is actually a decision tree, a tool that allows it, based on the information received, to select the most appropriate response among multiple possible solutions.
The intelligence of the robot is therefore limited to understanding the received message and choosing the appropriate response according to its decision tree. We are still far from an AI such as ChatGPT, with which open-ended conversations are possible.
For example, the best-known Voicebots are Siri for iPhone users and Alexa for Echo users.
To illustrate this evolution, let us dive into a sector that impacts every French citizen and where Voicebot technology is rapidly expanding : insurance.
Over the past three years, and especially since the Covid-19 pandemic, Voicebots have been widely deployed among insurance companies. Brands such as Axa, GMF, and Maaf, for example, have implemented one or several Voicebots.
The deployment of Voicebots has been driven in particular by the sharp increase in phone calls during the Covid-19 period, but also by the shortage of employees within call centers.
As an indication, a Voicebot deployed within a major insurance group can handle up to 400,000 calls per week.
From a customer perspective, the Voicebot provides an immediate response to their needs. Fast responses tend to improve customer satisfaction and therefore enhance the company’s image.
From an employee perspective, the Voicebot saves time and allows teams to focus on less repetitive and potentially more rewarding tasks. Employees are no longer trapped in repetitive operations and can dedicate more time to higher-value activities.
AI, robots, work… all the ingredients are there to address the controversial question. Yet another situation where the robot appears to replace humans, potentially threatening jobs.
That would certainly be true if robots were as capable as humans, but today this is not the case.
In reality, without a human team dedicated to monitoring and improving the robot, it would be only marginally effective and could even damage customer relationships.
A Voicebot does not naturally “speak human.”
Like a child learning to speak, it needs continuous exposure to new words in order to expand its vocabulary and better understand conversations. However, the conversations it is exposed to are often very repetitive. As a result, it repeatedly encounters the same vocabulary and may gradually see the richness of its language capabilities decline.
At that point, human intervention becomes essential to retrain the Voicebot by teaching or reminding it of new words and expressions.
Similarly, like a child, a Voicebot can misunderstand situations or misinterpret requests. Human intervention is then required to correct these interpretations.
Once again, the human role remains indispensable.
And unlike a child who eventually becomes an adult, a robot remains in a perpetual language-learning phase. It therefore requires continuous supervision throughout its entire lifecycle of use.
On top of that comes the monitoring required for any machine. Although a robot is capable of learning, it is still unable to repair itself in the event of bugs, or even to correct its own decision tree. Human involvement in its creation and maintenance therefore remains essential.
The Voicebot does not replace humans, it becomes a tool serving them.
On the one hand, the Voicebot requires attention and supervision, which must be provided by people. On the other hand, the robot can only handle standardized tasks falling within predefined processes. Human employees therefore remain indispensable for dealing with more complex or unusual situations.
That said, the Voicebot does perform tasks that were previously handled by call center employees. These employees, often occupying lower-skilled roles, may therefore appear to be replaced by automation.
But this is precisely where the concept of role evolution becomes meaningful.
By using Voicebots, companies free up employees’ time. This time can then be invested in upskilling these employees. Through the deployment of Voicebots, some insurance companies have increased training opportunities for these profiles, enabling them to evolve more rapidly and perform tasks that robots are not capable of handling.
This strategy has the advantage of increasing the number of qualified employees, particularly in customer-facing roles. It also contributes to improving overall service quality.
Furthermore, one could imagine these same employees, thanks to their operational field knowledge, being trained to supervise Voicebots and transition toward more technical Voicebot project management roles, positions that are often lacking within large organizations.
Voicebots eliminate certain jobs, but they also create new ones and provide employees with time to train and evolve.
It is therefore up to each company’s strategy to leverage Voicebots as a way to strengthen and develop internal capabilities.