When companies talk about artificial intelligence, discussions often focus on models, platforms or technology investments. However, the uses that really transform the daily lives of employees are sometimes much more modest.
At Deel, the first massively adopted automations did not involve large IT projects or vast transformation programs. They focused on preparing for customer meetings, writing reports after each call, sorting internal alerts and even documenting incidents.
“Taken individually, these are details. Put end to end, that’s 60 to 70% of the team’s time. We only discover them by observing actual work on the ground, never from a steering committee,” explains Anne-Lise Bouaziz-Klotz, Director of Customer Operations at Deel.
This observation sheds light on one of the most interesting phenomena in today’s adoption of artificial intelligence. The most transformative uses no longer necessarily arise in IT departments. They emerge as close as possible to operations, where employees accumulate hundreds of micro-frictions every day that are invisible to the rest of the organization.
Until AI, technological innovation followed a relatively stable path. The business departments formulated their needs, the IT departments selected the technologies, managed the projects then organized the deployments. Innovation was rare, expensive and largely centralized.
Generative artificial intelligence is changing this balance.
The end of a technological rarity
For a long time, the capacity to develop represented the main limiting factor. Companies could identify many automation opportunities without having the resources to make them happen. Each project required decisions, budgets and specialized teams.
Generative AI profoundly changes this equation. For the first time, non-technical employees can build tools themselves capable of meeting their daily needs. The difficulty no longer lies primarily in constructing the solution. It lies in identifying the problem that needs to be solved.
However, this knowledge is rarely found in technical teams; it belongs to employees confronted daily with operational irritants.
At Nabla, this dynamic appeared very early. The first uses did not emerge from a technological roadmap developed by an IT department. They appeared within go-to-market teams faced with an extremely concrete constraint: taking notes during customer meetings.
“The first concrete use case emerged from the go-to-market teams: taking notes during customer calls. The problem was simple: taking notes distracted attention, degraded the quality of listening to our customers and required devoting significant time to writing,” explains Delphine Groll, Co-Founder and Chief Operating Officer of Nabla.
The usage seems almost banal, and that is precisely what makes it interesting. Value does not come from technological prowess but from solving a daily problem identified by those who face it.
Once this first demonstration was carried out, the movement accelerated. “This use case served as an internal demonstration. Once the teams saw what AI could concretely solve, the field of possibilities opened up naturally,” continues Delphine Groll.
Business knowledge becomes a technological advantage
One of the paradoxes of artificial intelligence is that it increases the value of professional knowledge at the same time as it democratizes access to technology.
For several decades, the rare skill was mastery of technical tools. Today, this rarity is gradually moving towards a detailed understanding of the processes. Generative models already know how to write, synthesize, research or analyze. On the other hand, they do not know how to identify an organization’s irritants on their own or understand which tasks are unnecessarily consuming time.
This knowledge remains deeply rooted in the professions. This is precisely what Anne-Lise Bouaziz-Klotz observes at Deel. According to her, the most useful initiatives do not correspond to the large projects traditionally driven by technology functions.
“An IT department prioritizes large platform projects. We have automated the little invisible frictions of everyday life: the report after each customer call, the sorting of internal alerts, the documentation of bugs or the identification of the team to which to refer a problem. »
Artificial intelligence thus appears to be a technology particularly suited to optimizing real work, that which takes place between formal processes and which often remains invisible in organizational charts and IT roadmaps.
Support functions become the first AI laboratories
This development explains why some of the most advanced experiments are appearing today in departments rarely associated with technological innovation.
Human resources, customer support, operations and marketing are among the first beneficiaries of this new generation of tools. These functions rely largely on analysis, synthesis, information retrieval and coordination. So many tasks for which generative models provide immediate value.
At Deel, the first massively adopted uses concerned the automatic monitoring of payroll or risky onboarding, the preparation of customer meetings, the drafting of emails or even the generation of daily briefings intended for managers.
“Adoption was immediate because AI replaced hours of manual work,” summarizes Anne-Lise Bouaziz-Klotz.
At Nabla, the dynamic has also spread throughout the organization. Teams have gradually adopted automated note-taking tools, conversational assistants like ChatGPT or Claude, as well as different conversational intelligence solutions for sales teams. The AI functionalities integrated into Slack or Notion have been connected to the existing work environment in order to streamline access to information and certain daily tasks.
Innovation no longer only comes down from the technical teams to the business lines, but now goes up from the field to the organization.
A new generation of builders
This evolution is gradually transforming the very nature of the professions, so in Deel’s Customer Success teams, the role began to evolve as automation increased.
“Before, the Customer Success Manager spent his time executing: transmitting requests, following up, writing reports. Today, AI prepares the work, a draft email, a report, an enriched alert, and the human validates and decides. »
The change is significant, and the value focuses less on execution and more on judgment, customer relations and decision-making.
Even more interesting, teams gradually become responsible for their own automations.
“I build them, then I train them to bring them to life,” explains Anne-Lise Bouaziz-Klotz.
At Nabla, this development has even led to the appearance of new functions.
“We have created new roles such as those of prompt engineer or AI Lead,” explains Delphine Groll. The former design and evaluate the prompts used in the company’s products. The latter work directly with business teams to build agents that meet their operational needs.
Artificial intelligence is therefore no longer content with augmenting existing professions. It is already contributing to the emergence of new specialties within organizations.
The IT department changes role
This dynamic does not, however, mean a decline in IT departments. On the contrary, their importance could even increase. Simply, their role is evolving. Until now, they were responsible for building the systems, tomorrow they will be more responsible for their orchestration.
Security, data governance, regulatory compliance, architecture or access policies become even more critical issues as businesses become more autonomous.
The more agents and workflows people create, the more obvious the need for a common framework becomes. IT departments are gradually ceasing to be the sole producers of innovation to become the guarantors of its coherence.
This development is particularly visible in highly regulated sectors. “We operate in healthcare, a highly sensitive and regulated sector. This involves validation cycles by security teams, which can sometimes extend deployment times. This is not a brake in itself, it is a structural constraint that must be integrated,” underlines Delphine Groll.
A new geography of innovation
The testimonies collected from Nabla and Deel converge towards the same observation. The most relevant uses are not necessarily those imagined during major strategic planning exercises.
They often appear from daily frustrations, operational irritants or repetitive tasks identified by the teams themselves.
“The best ideas don’t come from a single team or leadership. They come from those who test, train and ask the right questions,” observes Delphine Groll.
Anne-Lise Bouaziz-Klotz formulates a similar idea when she explains that the best automations arise directly from the difficulties encountered in the field.
“Management provides the vision and unlocks the means. The business identifies the problem and builds the solution. »
This convergence is probably the most important signal. For several decades, businesses have viewed technology as a centralized capability. Artificial intelligence seems to introduce a different model, in which innovation first emerges from the professions before being structured at the scale of the organization.
The question is therefore no longer just what technologies to adopt. It becomes: how can we enable employees who know the problems best to build part of the solutions themselves?
For decades, digital innovation has flowed from the center to the periphery of the enterprise. Artificial intelligence seems to be taking the opposite path.
It is now born in the field, where the company’s daily problems are found, before being gradually industrialized at the scale of the organization.
