From Microsoft 365 Copilot to Copilot Studio: TotalEnergies at the forefront of AI transformation

A leading player in the energy sector, TotalEnergies stands out for its early and resolute commitment to Artificial Intelligence. As a trailblazer in adopting Machine Learning, the company took a pivotal step towards generative AI in 2023 by partnering with Microsoft.

30,000 Microsoft 365 Copilot licenses have since been rolled out to employees: just one year after deployment, 70% of staff were already recommending Copilot to their colleagues.

Building on this experience, the company is developing autonomous agents designed to support its strategic activities, including MARGE—a tool dedicated to industrial maintenance—and BuyerCompanion, an agent streamlining and simplifying certain procurement processes.

From natural gas and biofuels to electricity and oil, TotalEnergies is among the world’s leading energy producers and suppliers. Operating in nearly 120 countries, the company has developed comprehensive global expertise, fostered by a pioneering spirit nurtured over the past century.

This approach has resulted in significant progress in climate initiatives and energy transition, while establishing the group as a leader in emerging technologies such as AI. Artificial Intelligence now sits at the heart of its digital and multi-energy strategy, driven by the vision and conviction of CEO Patrick Pouyanné. The following is the perspective of Olivier Lacroze, Director of TotalEnergies Global Information Technology Services (TGITS).

The transformative power of technology

For many years, TotalEnergies has fully integrated AI into its operations, making it one of the first companies to use machine learning to optimize performance and efficiency. “AI is not merely a technology; it’s a strategic lever for reinventing our businesses,” emphasizes Olivier Lacroze.

In today's era of generative AI, this statement holds especially true, as it drives greater innovation and enhances the experiences of employees and customers alike. “AI accelerates our transformation; we leverage it to empower our staff, boosting their capacity for innovation and efficiency, all while keeping people at the center of decision-making,” Lacroze adds.

Deploying powerful tools at every level

In light of this, TotalEnergies’ leadership moved swiftly to embrace generative AI. The idea? To lead by example and drive change throughout the organization. "Our approach involves cultivating a robust AI culture by ensuring a broad base of employees are well acquainted with these technologies" Lacroze states. 

And adoption starts with experimentation. Equipping teams with the right tools is essential, which is why the company chose Microsoft 365 Copilot. Initially tested with a small group of employees at the end of 2023, the solution was then introduced to senior management in early 2024 and, within a few weeks, rolled out to almost 30,000 staff members.

Once again, TotalEnergies broke new ground, pioneering the large-scale deployment of generative AI to benefit its employees. “We quickly understood that implementing AI wasn’t enough if staff didn’t take ownership,” says Lacroze. “We put in place a comprehensive support system to embed these practices over the long term, before measuring and steering adoption.”

Building a strong AI culture

Simultaneously, the group developed a structured onboarding journey to foster a robust AI culture. In the initial weeks of implementing Microsoft 365 Copilot, they introduced an extensive, at-scale training program, organized hands-on workshops, and built a comprehensive support community.

To encourage adoption, TotalEnergies introduced a comprehensive awareness-raising and skills development program combining in-person training, webinars, Q&A sessions and dedicated resources. A dedicated team has tasked to oversee this strategy, respond to user needs, and provide ongoing support. In this context, an average of one event per day was held throughout the deployment period. 

The results were swift and tangible. Today, 70% of users recommend Copilot to their colleagues, reflecting both rapid adoption and perceived value. Usage has increased: 70% of employees use it weekly, and nearly half use it daily. These data demonstrate the seamless integration into daily operations and reflect its widespread adoption among staff members.

From acculturation to tangible results with AI agents

“We’re working on agents capable of searching for information, automating targeted tasks, and facilitating decision-making for frontline employees. These agents deliver real business value by addressing specific operational needs,” Lacroze explains enthusiastically. Driven by extensive acculturation, several high-value applications have been developed, particularly in the area of autonomous AI agents.

Among these is “MARGE”—short for Machine Assisted Reliability and Graving Engine—an agent based on Microsoft Copilot Studio and deployed with the support from Microsoft Industry Solutions teams. Designed for industrial maintenance, MARGE automatically analyses and classifies intervention reports. Processes that were previously completed manually now demonstrate enhanced reliability, evidenced by a 25% to 45% improvement in the identification of fault causes.

On the procurement side, the “BuyerCompanion” agent enables the industrialization of purchasing processes for transactions up to €50,000. Supported by Witivio, which offers AI solutions for Microsoft’s collaborative suite, and leveraging Microsoft Copilot Studio, Lacroze explains that “the agent streamlines processes ranging from drafting specifications to identifying suppliers.” In addition, it identifies relevant framework agreements based on the needs expressed, ensures compliance with the company's rules, and generates personalized recommendations. “This should allow us to save 10% on certain purchases.” Already operational in France, the agent could ultimately support 15,000 staff worldwide.

Innovation with complete control

In addition to MARGE and BuyerCompanion, over thirty applications have been deployed or are currently under development. The relevance of these applications is evaluated based on established criteria, including business value, usage frequency, and potential impact on company operations.

Subject to strict controls combining cybersecurity, IT architecture, and compliance, these agents have also prompted a rethink of internal organization and stakeholder roles. “TotalEnergies has set up an internal validation committee to assess the value an agent will deliver, its technical and architectural viability, cybersecurity, and to avoid redundant agents.”

Laying the foundations for sustainable AI adoption, step by step

“This is a pragmatic, incremental approach,” Lacroze adds. The roadmap for the coming months includes maintaining and scaling existing solutions, as well as developing multi-agent orchestration platforms. “TotalEnergies has tackled organizational challenges by clarifying responsibilities, shifting from a project-led to a product-led approach, establishing regular management rituals and proactively managing skills.”

It’s a new technological, cultural, organizational and human challenge—one the company is determined to meet. “We need to educate and reassure; it’s about augmenting people, not replacing them. There’s a balance to be struck between rigor and agility, and between innovation and security. That’s the key to deploying AI on a massive, responsible and sustainable scale,” concludes Olivier Lacroze.